sign in sign up
<<
Home , Finding aid, OCLC, Special collections

Making Archival Makingand Archival Accessible More Special Collections

Making Archival and Special Collections More Accessible OCLC Research

1504/215371, OCLC ISBN:-55653 1 3 -- 497 3 - -497 55653 - 1 978-

T: 1-800-848-5878 T: -6000 764 - +1-614 T: 764-6096 - F: +1-614 www..org/research Kilgour Place 6565 Kilgour -3395 43017 , www.oclc.org/research/themes/research-collections.html For more information about our work related to to related work our about information more For visit: support, please and collections research Making Archival and Special Collections More Accessible

Foreword by James Michalko © 2015 OCLC Online Center, Inc. This work is licensed under a Attribution 3.0 Unported License. http://creativecommons.org/licenses/by/3.0/

April 2015

OCLC Research Dublin, Ohio 43017 USA www.oclc.org/research

ISBN: 1-55653-497-3 (978-1-55653-497-3) OCLC Control Number: 906730124

Please direct correspondence to: James Michalko Vice President, OCLC Partnership [email protected]

Suggested citation: Making Archival and Special Collections More Accessible. 2015. Dublin, Ohio: OCLC Research. http://www.oclc.org/content/dam/research/publications/2015/oclcresearch- making-special-collections-accessible-2015-a4.pdf Acknowledgments

Special thanks to the following colleagues for their help with compiling and this report: Melissa Renspie, Linda Shepard and Eric Childress.

v

Contents

1 Foreword, by James Michalko...... 1

2 Taking Our Pulse: The OCLC Research Survey of Special Collections and – Executive Summary, by Jackie M. Dooley and Katherine Luce...... 5

3 Survey of Special Collections and Archives in the United Kingdom and Ireland – Summary and Recommendations, by Jackie M. Dooley, Rachel Beckett, Alison Cullingford, Katie Sambrook, Chris Sheppard, and Sue Worrall...... 11

4 Taking Stock and Making Hay: Archival Collections Assessment, by Martha O’Hara Conway and Merrilee Proffitt...... 17

5 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation, by Michele Combs, Mark A. Matienzo, Merrilee Proffitt, and Lisa Spiro...... 39

6 Thresholds for Discovery: EAD Analysis in ArchiveGrid, and Implications for Discovery Systems, by Marc Bron, Merrilee Proffitt and Bruce Washburn...... 63

7 The IS the Interface: Better Description for Better Discovery of Archives and Special Collections, by Jennifer Schaffner...... 85

8 “Capture and Release”: Digital Cameras in the Reading Room, by Lisa Miller, Steven K. Galbraith, and the RLG Partnership Working Group on Streamlining Photography and Scanning...... 99

9 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections, by Dennis Massie...... 113

Foreward 1 Foreword

James Michalko Vice President, OCLC Research Library Partnership

1 Making Archival and Special Collections More Accessible

“, archives, and cultural institutions hold millions of items that have never been adequately described. These items are all but unknown to, and unused by, the scholars those organizations aim to serve. …Nationally, this represents a staggering volume of items of potentially substantive intellectual value that are unknown and inaccessible to scholars.”

—from the CLIR announcement of the Cataloging Hidden Special Collections and Archives program, 17 March 2008.

“What is important about and serials is that moving digital surrogates and newly produced works to the network level generates aggregations operating at a scale that advances existing lines of inquiry and opens new ones and makes scholars and students more productive, even when using individual works. These same criteria must form the heart of the value proposition for special collections.”

—Donald J. Waters in “The Changing Role of Special Collections in Scholarly Communications.” Research Library Issues (A bimonthly report from ARL, CNI, and SPARC), no. 267 (Dec. 2009): 35-36.

2 Foreward

OCLC Research supports research institutions summaries of each are included in this volume in collaboratively designing their future. This and include recommendations for action that leads us to work in areas that will reduce are notable for their good sense and moderated redundant efforts, change community economics, scope, and will have significant impact. The transform processes and respond to known and counts, charts and associated analysis that were emerging needs. Revealing the hidden assets the heart of this effort are available from the OCLC stewarded by research institutions so they Research website. They continue to be among can be made available for research is a prime the most referenced of our work in this area. opportunity for creating and delivering new Understanding the scope and range of value. Over the last seven years we have worked unprocessed and therefore hidden collections to support change in the end-to-end process made it clear that the community needed that results in archival and special collections renewed encouragement to take on the daunting materials being delivered to interested users. task of describing these collections. In Taking The overarching goal of this work continues to be Stock and Making Hay: Archival Collections the achievement of economies and efficiencies Assessment, we urged institutions to undertake that permit these materials to be effectively an accurate census of their archival collections as described, properly disclosed, successfully a foundation for acting strategically in meeting discovered and appropriately delivered. user needs, allocating available resources, Achieving control over these collections in and securing additional funding. This kind an economic fashion will mean that current of data about collections informs important resources can have a broader impact or be decisions regarding management, invested elsewhere in other activities. processing priorities, and selection and other activities associated with digitization Special collections and archives are part of and exhibit preparation. Understanding that almost every library collection no matter the resources for this type of work are scarce size or type of library. They are ubiquitous at and the scale can be daunting, this report research institutions and present a similar provides a variety of practical approaches that schizophrenic management challenge across encourage progress by avoiding prescriptions. the community. They are treasures. They are burdens. They are valued. They are costly. Of course, understanding the scope of local They are important. They are hidden. collections does not mean that they are properly described, which is a necessary condition for them I have frequently framed the goal of the concerted to be discovered. For more than a decade the attention that OCLC Research has given to target for describing archival collections has been archives and special collections as a change in the the Encoded Archival Description (EAD) standard, cost/benefit equation associated with these types which has long been considered a high value but of materials. There are two ways to increase the costly descriptive mechanism. Consequently it has ratio. You can affect the denominator by reducing been implemented by a bare majority of archives, the ongoing investment necessary to steward and many institutions have been daunted by its these materials, or you can change the numerator related political, logistical and technical issues. In by increasing the utility of the materials for Over, Under, Around and Through: Getting Around teaching and research both locally and globally. Barriers to EAD Implementation we offered tools— Our work has tried do both. We captured that information, persuasion and technology—to intent in the phrase “Mobilizing Unique Materials.” help practitioners surmount these roadblocks. As is often the case with OCLC Research we began This enormously practical document encourages our work in this area with a system-wide view by providing lots of examples and offering that described the scope of the opportunity and simplification of the unnecessarily complex. the problem. Our two surveys—Taking Our Pulse: The OCLC Research Survey of Special Collections and Archives in the US and Canada and the Survey of Special Collections and Archives in the UK and Ireland—were large multi-year efforts that brought the community an updated understanding of the state of these types of materials. The executive

3 Making Archival and Special Collections More Accessible

This complexity grows over time as standards Nor can they be valued if they only stay home. expand, practice changes, and local choices In Tiers for Fears: Sensible, Streamlined Sharing drive the descriptive effort. How well the of Special Collections we argue that interlending resulting descriptions actually serve to advance of actual physical items from special collections the discovery of materials has not been much for research purposes should be supported. studied in the literature. Certainly it has not been Special collections have long done this for rigorously demonstrated or seriously challenged. exhibitions but providing the physical item In Thresholds for Discovery: EAD Tag Analysis in to the distant scholar is rare and elicits a fear ArchiveGrid, and Implications for Discovery Systems response in many special collection managers. we add to the evidence base by examining This exhortatory piece acknowledges that the large (120,000+) corpus of EAD documents trust is essential to establishing new lending harvested and made available through the practices and helps practitioners evaluate the ArchiveGrid aggregation maintained by OCLC tiers of effort required to lend certain materials Research. Our analysis looks to determine how and the trustworthiness of the other parties well the documents support various aspects to the transaction. I’m pleased to say that this of online discovery. And we conclude that the piece has changed practice at an impressive picture for archival discovery and EAD is decidedly cadre of world-class special collections. mixed. We also offer advice about where it would This volume stands as evidence of a body of be worthwhile to invest local descriptive effort effort devoted to areas that we think hold in elements particularly crucial to discovery. enormous promise for enhancing the library’s Understanding the way tags are used in systems value proposition. The unique materials that support search, browse, results displays, stewarded by our institutions need to release sorting and limiting is one very important way their value to a global audience of researchers to evaluate the investment in description. It in ways that will enhance the reputation of the proceeds, however, from a system vantage. steward. That will happen only when we devote Guiding this investment on the basis of what structured effort to the full range of selection, users need and value is the intent of the description, discovery and delivery. With the synthesis of user studies that we provided in right attention as signaled by the pieces in this The Metadata IS the Interface: Better Description volume we can mobilize our unique materials for for Better Discovery of Archives and Special maximum value in the networked environment. Collections. This well organized, definitive survey concludes that we best respond to users by putting the right descriptive metadata in the right places. Those places are network discovery environments not local portals. Finally the volume concludes with two very influential pieces that urge practitioners to go beyond traditional bounded practices in order to satisfy their users who have discovered special collections and archival materials. In “Capture and Release”: Digital Cameras in the Reading Room we acknowledge the ubiquity of digital cameras and other mobile capture devices which has led researchers to expect to use cameras in reading rooms. We argue that embracing and supporting the use of these devices provides benefits to researchers, repositories, and collection materials. We provide advice to support changes in practice that will satisfy on-site researcher expectations and are consistent with institutional practice. Special collections cannot release their value if they are camera-shy.

4 Taking Our Pulse: The OCLC Research Survey of Special Collections and Archives 2 Taking Our Pulse: The OCLC Research Survey of Special Collections and Archives

Executive Summary

Jackie M. Dooley Program Officer

Katherine Luce Research Intern

OCLC Research

This executive summary was originally published in October 2010 by OCLC Research in Taking Our Pulse, The OCLC Research Survey of Special Collections and Archives at http://www.oclc.org/content/dam/research/publications/library/2010/2010-11.pdf.

5 Making Archival and Special Collections More Accessible

Executive Summary

Special collections and archives are increasingly The top three “most challenging seen as elements of distinction that serve to issues” in managing special collections differentiate an academic or research library were space (105 respondents), born- from its peers. In recognition of this, the digital materials, and digitization. Association of Research Libraries conducted • We updated ARL’s survey instrument a survey in 1998 that was transformative and extended the subject population to and led directly to many high-profile encompass the 275 libraries in the following initiatives to “expose hidden collections.” five overlapping membership organizations: As this OCLC Research report reveals, • Association of Research Libraries however, much rare and unique material (124 universities and others) remains undiscoverable, and monetary resources are shrinking at the same time • Canadian Academic and Research that user demand is growing. The balance Libraries (30 universities and others) sheet is both encouraging and sobering: • Independent Research Libraries Association • The size of ARL collections has grown (19 private research libraries) dramatically, up to 300% for some formats • Oberlin Group (80 liberal arts colleges) • Use of all types of material has • RLG Partnership, U.S. and Canadian increased across the board members (85 research institutions) • Half of archival collections The rate of response was 61% (169 responses). have no online presence • While many backlogs have decreased, almost as many continue to grow Key Findings • User demand for digitized A core goal of this research is to incite change collections remains insatiable to transform special collections, and we have threaded recommended actions throughout this • Management of born-digital archival section. We focused on issues that warrant shared materials is still in its infancy action, but individual institutions could take • Staffing is generally stable, but immediate steps locally. Regardless, responsibility has grown for digital services for accomplishing change must necessarily be distributed. All concerned must take ownership. • 75% of general library budgets have been reduced Assessment • The current tough economy renders “business as usual” impossible A lack of established metrics limits , analyzing, and comparing statistics across the special collections community. Norms for tracking and assessing user services, metadata creation, , digital production, and other activities are necessary for measuring institutions against community norms and for demonstrating locally that primary constituencies are being well served.

6 Taking Our Pulse: The OCLC Research Survey of Special Collections and Archives

ACTION: Develop and promulgate The preservation needs of audiovisual collections metrics that enable standardized (both audio and moving image) are well known measurement of key aspects of special to be staggering, and our data confirm that these collections use and management. materials have by far the most serious problems.

ACTION: Take collective action to share resources for cost-effectivepreservation Collections of at-risk audiovisual materials. ARL collections have grown dramatically since 1998, ranging from a 50% increase in the mean for printed volumes and archival collections to 300% for visual and moving-image materials. User Services Two thirds of respondents have special collections in secondary storage. As general print More than 60% of respondents stated that use by collections stabilize, such as through shared faculty, undergraduates, and visiting researchers print initiatives and digital publication, a need has increased over the past decade. Nearly half, for more stacks space for special collections however, were unable to categorize their users by will become all the more conspicuous. The type, even those in their primary user population. arguments to justify it will have to be powerful. User services policies are evolving in positive The amount of born-digital archival material ways: most institutions permit use of digital reported by respondents is minuscule relative cameras and 90% allow access to materials to the extant content of permanent value: in backlogs. More than one third send original the mean collection size is 1.5 terabytes, the printed volumes on , while median a mere 90 gigabytes. It is striking that nearly half supply reproductions. Conservative only two institutions hold half of the material vetting of requests may, however, result in reported, and only thirteen hold 93% of it. unwarranted denial of all three types of access. Receipt of a gift is the most frequently stated impetus for undertaking a new collecting ACTION: Develop and liberally implement emphasis. Some respondents noted, however, exemplary policies to facilitate rather that they do not plan to acquire other materials than inhibit access to and interlibrary to strengthen the new area, which may signal that loan of rare and unique materials. the gift was outside the library’s areas of strength or need. Such gifts sometimes become a liability over time. of unwanted materials, some of which have languished unprocessed Cataloging and Metadata for years, occurs for appropriate reasons but is not widely practiced. Informal collaborative The extent to which materials appear in online collecting is fairly widespread on a regional basis, catalogs varies widely by format: 85% of printed but formal arrangements of any kind are rare. volumes, 50% of archival materials, 42% of maps, and 25% of visual materials are accessible online. Relative to ARL’s 1998 data, 12% more ACTION: Identify barriers that limit printed volumes have an online record, as do collaborative . 15% more archival materials and 6% more Define key characteristics and desired maps. This limited progress may be attributable outcomes of effective collaboration. in part to lack of sustainable, widely replicable methodologies to improve efficiencies.

ACTION: Compile, disseminate, and adopt a slate of replicable, sustainable methodologies for cataloging and processing to facilitate exposure of materials that remain hidden and stop the growth of backlogs. 7 Making Archival and Special Collections More Accessible

ACTION: Develop shared capacities to create The institutional archives reports to the metadata for published materials such library in 87% of institutions, while two thirds as maps and printed graphics for which have responsibility for records management cataloging resources appear to be scarce. (of active business records). The challenges specific to these materials should therefore be core concerns of most libraries—and it is On the other hand, great strides have been in this context that the impact of born-digital made with archival finding aids: 52% of ARL content is currently the most pervasive. collection guides are now accessible online, up from 16% in 1998. Across the entire population Digitization the figure is 44%, which would increase to Nearly all respondents have completed at least 74% if all extant finding aids available locally one special collections digitization project and/ were converted. The other 26% reveals the or have an active digitization program for special archival processing backlogs that remain. collections. One fourth have no active program, and the same number can undertake projects ACTION: Convert legacy finding aids using only with special funding. More than one third affordable methodologies to enable Internet state that they have done large-scale digitization access. Resist the urge to upgrade or of special collections, which we defined as a expand the data. Develop tools to facilitate systematic effort to digitize complete collections— conversion from local databases. rather than being selective at the item level, as has been the norm—using production methods that are as streamlined as possible. Subsequent follow-up with respondents has Backlogs of printed volumes have decreased at revealed, however, that the quantities of material more than half of institutions, while one fourth digitized and/or production levels achieved have increased. For materials in other formats, generally were not impressive or scalable. increases and decreases are roughly equal.

Archival ACTION: Develop models for large- scale digitization of special collections, The progress made in backlog reduction for including methodologies for selection archival materials is aided by the fact that 75% of of appropriate collections, security, safe respondents are using minimal-level processing handling, sustainable metadata creation, techniques, either some or all of the time. Tools for and ambitious productivity levels. creation of finding aids have not, however, been standardized; some institutions use four or more. One quarter of responding institutions have licensing contracts with commercial vendors to digitize materials and sell access. It would be useful to learn more about the existing corpus of digitized materials, particularly rare books, some important collections of which are not available via open-access repositories.

ACTION: Determine the scope of the existing corpus of digitized rare books, differentiating those available as from those that are licensed. Identify the most important gaps and implement collaborative projects to complete the corpus.

8 Taking Our Pulse: The OCLC Research Survey of Special Collections and Archives

Born-Digital Archival Materials The areas most often mentioned in which education or training are needed to fulfill The data clearly reveal a widespread lack of basic the institution’s needs were born-digital infrastructure for collecting and managing born- materials (83%), information technology digital materials: more than two thirds cited lack (65%), intellectual property (56%), and of funding as an impediment, while more than cataloging and metadata (51%). half noted lack of both expertise and time for planning. As a result, many institutions do not even know what they have, access and metadata ACTION: Confirm high-priority areas in which are limited, only half of institutions have assigned education and training opportunities are responsibility for managing this content, few have not adequate for particular segments of the collected more than a handful of formats, and professional community. Exert pressure on virtually none have collected at scale. Clearly, appropriate organizations to fill the gaps. this activity has yet to receive priority attention due to its cost and complexity. Community action could help break the logjam in several ways. The gradual trend in recent decades toward integration of once-separate special collections ACTION: Define the characteristics of continues; 20% of respondents have done born-digital materials that warrant their this within the past decade. Multiple units management as “special collections.” continue to exist at one in four institutions.

ACTION: Define a reasonable set of basic steps for initiating an institutional program for responsibly managing born-digital archival materials.

ACTION: Develop use cases and cost models for selection, management, and preservation of born-digital archival materials.

Staffing The norm is no change in staff size except for in technology and digital services, which increased at nearly half of institutions. Even though more than 60% of respondents reported increased use of collections, staffing decreased in public services more frequently (23%) than any other area. Across the population, 9% of permanent special collections staff are likely to retire within the next five years.

9 Making Archival and Special Collections More Accessible

10 Survey of Special Collections and Archives in the United Kingdom and Ireland 3 Survey of Special Collections and Archives in the United Kingdom and Ireland

Summary and Recommendations

Jackie M. Dooley King’s College OCLC Research Chris Sheppard Rachel Beckett University of Leeds University of Manchester Sue Worrall Alison Cullingford University of Birmingham Bradford University A co-publication of OCLC Katie Sambrook Research and RLUK

This summary and recommendations was originally published in February 2013 by OCLC Research at http://www.oclc.org/content/dam/research/publications/library/2013/2013-01-sumrecs.pdf.

Read the complete Survey of Special Collections and Archives in the United Kingdom and Ireland report at http://www.oclc.org/content/dam/research/publications/library/2013/2013-01.pdf.

11 Making Archival and Special Collections More Accessible

Executive Summary

It has become widely recognised across the We asked respondents to name their three academic and research libraries sector that ‘most challenging issues.’ The following special collections and archives play a key role were the most frequently cited: in differentiating each institution from its peers. • Outreach (broadly defined) In recognition of this, Research Libraries UK (RLUK) established the workstrand ‘Unique and • Space and facilities (particularly Distinctive Collections’ in support of its strategic for collections) aims for 2011-2014. The UDC workstrand will • Born-digital materials identify ways in which special collections can ‘make the most of their potential for research, • Collection care teaching and community engagement.’ This • Cataloguing and archival processing survey forms part of the overall project by gathering data to enable better understanding One hundred twenty-two academic and of the sector. It was conducted as a collaboration research libraries with significant special between RLUK and OCLC Research. collections received invitations to participate in the survey. The rate of response was 67% (82 As this report reveals, we face numerous responses), including 100% of RLUK members. challenges if we are to maximise potential and bring special collections to the This report presents a summary and analysis attention of those whose research or of the data for all respondents, for RLUK learning would benefit from their use. members, and for non-RLUK respondents, with a complete set of data figures and tables for A few of the most salient issues each. Also included is a comparison of the RLUK that emerged from the data: data with that of the Association of Research • Alignment of special collections with Libraries (US) members who responded to an institutional missions and priorities OCLC Research survey of the is an ongoing challenge. and Canada (Dooley and Luce, 2010). • The special collections sector is undergoing a major culture shift that Key Findings mandates significant retraining and careful examination of priorities. Outreach and User Services • Philanthropic support is limited, as More than half of respondents stated that use are ’ fundraising skills. of special collections by all types of users has • Use of all types of material has increased over the past decade. Few, however, increased across the board. were able to categorise their users by type, even those in their primary user population: 90% of • Users expect everything in libraries users were reported as ‘other’ (i.e., type of user and archives to be digitised; national not identified). This could be problematic if it strategies for digitisation of rare and results in an inability to demonstrate the extent unique materials are therefore needed. to which the primary audience is being served. • Many cataloguing backlogs have decreased, User services policies are evolving in productive while some continue to grow. ways: three-quarters of institutions permit • One-third of archival collections are not use of digital cameras, and up to 80% allow discoverable in online catalogues. access to printed volumes and archival materials in backlogs. On the other hand, • Management of born-digital archival 81% do not permit interlibrary loan, even of materials remains in its infancy; upper reproductions, which could be considered management must actively support this a disservice to distant researchers. important work to ensure progress. Despite these very promising data, many respondents indicated that the need to embrace 12 Survey of Special Collections and Archives in the United Kingdom and Ireland new modes of outreach and service presents recent years. Although informal collaborative enormous challenges. This appears to stem collecting is fairly widespread on a regional basis, from two principal factors: staff skills are being formal arrangements of any kind are rare. stretched by the need to undertake new duties, and, as a result, fulfillment of ‘traditional’ Born-digital Materials responsibilities is thereby rendered more difficult. The data clearly reveal a widespread lack of basic infrastructure for collecting and managing Staffing born-digital materials. Sixty percent cited lack As mentioned above, the need to undertake of funding as an impediment, while only slightly new duties is proving to be a major challenge. fewer noted lack of both expertise and time The areas most often mentioned in which for planning. As a result, many institutions do education or training are needed to fulfill the not even know what they have, access and institution’s needs were born-digital materials, metadata are limited, half of institutions have fundraising, intellectual property, and outreach. not yet assigned responsibility for managing this content, few have collected more than a The data show that the mean number of handful of digital formats, and virtually none permanent special collections staff across the have collected at the level that is warranted. entire population is 16.6 FTE. The median is only six, which reveals wide variation across The amount of born-digital archival material institutions. This comparison is very different reported is minuscule relative to the extant when the data are analyzed by type of institution. content that warrants being preserved in Forty percent of respondents have experienced archives: the mean collection size is only 2,800 an increase in the number of professionally gigabytes, and the median is zero. It is striking qualified staff in recent years, while 29% that only two institutions hold 80% of the material had an increase in support staff. Across the reported, while five hold 99%. Clearly, academic population, 7% of special collections staff are and research libraries have barely scratched likely to retire within the next five years. the surface of the born-digital challenge. A trend exists toward integration of once- separate special collections departments— Digitisation more than half of respondents have Perceived pressure to digitise collections done so within the past decade. comprehensively seems to be ubiquitous. Ninety- seven percent of respondents have completed at Collections least one special collections digitisation project and/or have an active digitisation programme that Insufficient space for collections, or inadequate includes special collections. Progress is impeded, space needing renovation to satisfy current however, by the fact that less than half can needs, ranked very high among the ‘challenging undertake projects without special funding, while issues.’ More than one-third of respondents one-third have a recurring budget for digitisation. have special collections in secondary storage. Deaccessioning of unwanted materials, some One-third stated that they have done large- of which have not been processed many years scale digitisation of special collections (defined after they were acquired, occurs for appropriate as a systematic effort to digitise complete reasons but is practiced by only a few. Review collections and employing production methods of unprocessed collections for retention could that are as streamlined as possible) rather than be one way to contend with insufficient space. selecting and interpreting particular items. As the size of general print collections stabilise, More than 40% have licensing contracts such as through shared print initiatives with commercial vendors to digitise and digital publication, a need to add more materials and sell access. storage space for special collections would become all the more conspicuous. Archival Collections One-third of respondents have undertaken While shared archival online catalogues have one or more new collecting emphases in proven to be successful discovery hubs, only

13 Making Archival and Special Collections More Accessible one half of archival finding aids are accessible Collection care online. This percentage would increase to The preservation needs of both audiovisual 82% if all extant finding aids available only and born-digital materials are well known at the host institution were converted. The to be huge, and our data confirm this. remaining 18% (no finding aid exists) reveals the archival processing backlogs that remain. The most widespread collection care problems The progress made in backlog reduction may be are conservation repair of materials to enable due, at least in part, to the use of minimal-level their use and rehousing into improved boxes processing techniques by 70% of respondents. and other housings. Issues related to quality of storage facilities were cited by about 40%. The institutional archives reports to the library in two-thirds of institutions, while nearly half have responsibility for records management Metrics (of active business records). The challenges A lack of established metrics placed some specific to these materials should therefore constraints on the data that respondents be core concerns of most libraries—and it is could contribute and our ability to analyse in this context that the impact of born-digital it closely. Norms for tracking and assessing content is currently the most pervasive. user services, metadata creation, archival processing, digital production, and other Cataloguing and Metadata activities would make it more feasible to establish reliable community norms against Backlogs of printed volumes have decreased which to measure individual institutions. at nearly half of institutions, while somewhat fewer backlogs have increased. For materials We did not explore the particular purposes in other formats, increases and decreases that would be served by deployment of a set are roughly equal. The continuing existence of uniform metrics; it would be important of backlogs may be attributable in part to to do so before undertaking such work. the lack of sustainable, widely replicable methodologies to improve efficiencies. The extent to which materials appear in online catalogues varies widely by format: 78% of printed volumes, 64% of archival materials, half of maps, and one-third of visual materials are accessible online.

14 Survey of Special Collections and Archives in the United Kingdom and Ireland

Recommendations These recommendations were formulated by the authors of this report and are wholly based in analysis of the survey data. Participants in the RLUK Unique and Distinctive Collections symposium held at the University of Aberdeen on 29 March 2012 very usefully vetted an early version, which the authors then significantly revised. Note: This is not a set of recommendations officially endorsed by RLUK or intended for RLUK action; a forthcoming report on the UDC workstrand will fulfill that need. In general, under each category we consider the first recommendation a higher priority than the other(s) in that group (e.g., we feel that 1.1 would potentially have a higher impact than 1.2). 1. Staffing 5. Digitisation 1.1. Analyse the array of duties performed 5.1. Develop both a national strategy by special collections staff and identify for continued digitisation of special the new skills and expertise needed collections and a national gateway for to move the profession forward in discovery of digitised content. As part alignment with institutional missions. of the strategy, identify sustainable funding strategies and international 1.2. Develop a plan to provide educational and partners with which to collaborate. development opportunities in areas, both traditional and emergent, in which skills 5.2. Develop cost-effective models for need enhancement across the sector. large-scale digitisation of special collections that take into account the 2. External Funding special needs of these materials while 2.1. Develop a set of arguments to assist also achieving high productivity. institutions with development 6. Archival Collections of external sources of funding in support of special collections. 6.1. Convert print archival catalogues using affordable methodologies to enable 3. User Services Internet access. Develop approaches 3.1. Develop an outreach toolkit, including to modifying existing descriptions that case studies illustrating best practices, to strike a balance between incurring build skills for presentation, promotion, overheads and being effective for and engagement with special collections. discovery. Develop tools to facilitate 3.2. Develop pricing models, templates, conversion from local databases. and shared policies for user-initiated 6.2. Develop a shared understanding of digital scanning to encourage the goals, characteristics, and benefits consistency across the sector. of ‘simplified archival processing.’ 4. Born-digital Materials 6.3. Establish a methodology to assess 4.1. Define the basic steps involved in initiating unprocessed archival collections and a program for managing born-digital develop a plan to make the national archival materials to assist libraries collection more fully accessible. that have not yet begun this work. 7. Metrics 4.2. Investigate the feasibility of extending 7.1. Determine the potential value and broadly across the sector the adoption uses of metrics for reporting core of successful technical environments statistics (e.g., collection size, users, for managing born-digital materials outreach efforts, catalogue records) that have been developed by a across the sector. If warranted, define small number of UK institutions. categories and methodologies and encourage their use across the sector.

15 Making Archival and Special Collections More Accessible

8. Collection Development 8.1. Define key characteristics and desired outcomes of meaningful collaborative collection development, and encourage collaborations in areas of national significance. 8.2. Scrutinise local collecting policies to determine how well they reflect the institutional mission and can feasibly be implemented. 9. Cataloguing and Metadata 9.1. Collaborate to share expertise and create metadata for cartographic materials to enable improved discovery of the national collection. 9.2. Build on the findings of RLUK’s ‘hidden collections’ survey of print materials to identify national cataloguing priorities. 10. Collection Care 10.1. Further inflect the COPAC collection management tool to meet the requirements of special collections. Investigate its potential for determining priorities for preservation and other management activities across the national print collection. 10.2. Take collective action to share resources for cost-effective preservation of at-risk audiovisual and born-digital archival materials. 11. Building community 11.1.Identify beneficial ways in which to build productive relationships across the diverse community of special collections libraries that participated in this survey.

16 Making Archival and Special Collections More Accessible 4 Taking Stock and Making Hay: Archival Collections Assessment

Martha O’Hara Conway Associate Director, Special Collections Library,

Merrilee Proffitt Senior Program Officer OCLC Research

This paper was originally published in September 2011 by OCLC Research at http://www.oclc.org/content/dam/research/publications/library/2011/2011-07.pdf .

17 Making Archival and Special Collections More Accessible

Acknowledgments What is Archival Collections This report would not have been possible without Assessment? the support of a working group from the OCLC Research Library Partnership. Members of that In this report, the term “archival collections group included: Martha O’Hara Conway, University assessment” is used to refer to the systematic, of Michigan; de Lorenzo, University of purposeful gathering of information about , Berkeley; Christine Di Bella, Institute of archival collections. It includes collection Advanced Study; Sarah Stauderman, Smithsonian surveys of all kinds, including those undertaken Institution. Members represented a broad range for purposes of appraisal, setting processing of knowledge and perspectives, and their real and other priorities, conservation decision- world, hands on experiences with collections making, and collection management. assessment was invaluable in writing this report. An accurate census of its archival collections Many thanks to colleagues who read an early enables the institution to act strategically draft and provided helpful suggestions and in meeting user needs, allocating available recommendations: Jackie Dooley, Ricky Erway, resources, and securing additional funding. Jennifer Schaffner (OCLC Research), Rachel Onuf The systematic gathering of quantitative and (consultant) and Mark Allen Greene (American qualitative data about collections makes Heritage Center). Thanks are also due to John possible the creation of adequate, consistent, Fleckner (Smithsonian Institution), Barclay collection-level descriptions; affords a better Ogden (, Berkeley), understanding of unmet preservation needs; Mary Ide (WGBH), Nancy Elkington (OCLC and informs important decisions regarding Research), and David Molke-Hansen (retired, collection management, processing priorities, Historical Society of Pennsylvania) for giving and selection and other activities associated background interviews on their involvement with digitization and exhibit preparation. with early collections assessment projects. A Common Approach? Introduction Although a number of institutions have Archival collections assessment is an important undertaken collections assessments, a single, component of a successful collections commonly-understood approach neither exists management program. In most institutions, nor is practical. Rather than recommending however, conducting an assessment is feasible a single strategy or advocating a particular only with additional resources. For this and a approach, this report identifies and characterizes number of other reasons, collections assessment existing surveys that can be used as-is by, or serve has not been a regular part of collections as models for, librarians, , and others management practice. In recent years, however, who are considering collections assessment a number of institutions have created or adapted to meet one or several institutional needs. collections assessment tools, employed them It describes the many possible components successfully, and made them available for use of collections assessment; emphasizes the by others. The wheel has been invented. importance of approaching collections assessment with an informed understanding of its purpose and desired outcome(s); and provides pointers to existing methodologies and tools that have been used by various institutions.

18 Taking Stock and Making Hay: Archival Collections Assessment

Our goal in assembling this report and making More Product, Less Process it available to the widest possible audience is to encourage the archival and special collections Cataloging and processing backlogs have long communities to use existing approaches in been the bane of the cultural heritage institution, order to leverage good work, foster a growing and calls for addressing them have been around community of practice, and encourage efficiencies for almost as long as the backlogs themselves. for institutions both individually and collectively. One of the most recent of these was put forth in “More Product, Less Process: Revamping Traditional Archival Processing,” an article by How to Use This Report Mark Greene and Dennis Meissner that gave voice to the small but growing number of archivists who This report provides both food for thought have quietly abandoned traditional approaches and fuel for activity. It presents a rationale to archival processing in favor of those that for conducting a collections assessment; expedite user access to archival collections (2005). describes the components of archival collections assessment; and encourages readers to In their article, Greene and Meissner issue a consider their own needs and capacities. call for change that specifically references the Additionally, we hope this report will serve to successful reduction of cataloging backlogs inspire and empower those who are considering in large research libraries through various collections assessment by suggesting an array of procedural and technical and possibilities that can be readily applied to meet by redefining quality. In redefining quality as immediate and/or long-term needs. Appendix it applies to processing, they assert that, A contains pointers to a variety of exemplar it must be our aim to provide sufficient projects, many of which have tools and more physical and intellectual access to collections information available online. Appendices B and for research to be possible, without the C are links to project documentation, which necessity of processing each collection to an contain useful instructions and definitions. ideal or arbitrary standard. We should be paying more attention to achieving basic physical and intellectual control over, and Current Context: Tackling thus affording research access to, all our holdings, rather than being content to process the Backlog Problem a few of them to perfection. What this means It is no longer a “dirty little secret” (Tabb is that all collections should have collection- 2004, 123) that libraries, archives, and cultural level intellectual control before any collection institutions hold significant amounts of receives folder-level intellectual control. More special collections material that have not been importantly, researchers cannot come to do adequately described and therefore are not research if at least minimal information about known, cannot be discovered, and will not be the collections is not available to them. (237) used. These uncataloged and unprocessed (i.e., In other words, “Describe everything in “hidden”) collections have become the focus general before describing anything in detail.” of considerable attention in recent years and And make those descriptions available efforts to address the problems they represent to the widest possible audience. are numerous, varied and well documented (ARL 2008; CLIR 2011; Hewitt and Panitch 2003; Pritchard 2009; Schreyer 2007; Steele 2008).

19 Making Archival and Special Collections More Accessible

Exposing Hidden Collections problem. Fifteen percent of printed volumes still are not cataloged online, while for other formats, It is worth noting that the same principle was the situation may even have worsened. The survey endorsed by those participating in the “Exposing suggests that a large percentage of materials Hidden Collections” conference that took place lack online access, including 44% of archives and in September 2003 at the Library of Congress. A , 58% of cartographic materials, and working conference planned by the Association almost 25% of video and audiovisual materials. of Research Libraries (ARL) Task Force on Special Perhaps the most sobering statistic: 71% of Collections, “Exposing Hidden Collections” born digital materials held in special collections served as a forum for interaction between various are not represented in online catalogs. (Dooley communities of professionals and set the stage and Luce 2010, 46). This represents a staggering for the collaborative development of an action amount of material that is neither known by plan aimed at surfacing “hidden” collections. One nor available to the research community. of the overriding themes of the conference was “Some access to all is preferable to no access to some.” In fact it was proposed at the outset of the Help Is on the Way two-day event that one of the outcomes of the Heightened awareness of both the scope and the conference should be “a pledge by participants implications of the hidden collections “problem” to return to their institutions committed to is, fortunately, matched by a number of new providing a web-accessible collection-level record and existing initiatives aimed at addressing it. for all unprocessed materials (ARL 2009).” The most recent of these—launched in 2008— This apparently has proven to be either more is the Cataloging Hidden Special Collections difficult or more problematic than the conference and Archives Program. With generous funding participants imagined. For some institutions, from The Andrew W. Mellon Foundation, providing a web-accessible collection-level the Council on Library and Information record for an unprocessed collection is difficult Resources (CLIR) administers this national because an appropriate record neither exists program that awards grants in support of nor is easily created until the collection is “innovative, efficient description of large processed. Some institutions might be reluctant volumes of material of high value to scholars.” to provide a web-accessible collection-level Projects are evaluated and selected for record for an unprocessed collection because funding according to the following criteria: it would suggest that the collection is available for use when, for any of a variety of reasons, it • potential national impact on might not be. Although few of the conference scholarship and teaching; participants followed through on that pledge, • use of innovative and/or highly efficient some institutions have made descriptions of their approaches to description that could unprocessed collections available via the web. serve as models for others; • adoption of workflow and outreach practices Cataloging Hidden Collections that maximize connections to scholarly A 1998 survey of ARL member institutions and other user communities; and revealed that “significant portions” of special • application of descriptive and other collections material have not been processed or standards that would provide interoperability cataloged and therefore are not known, cannot and long-term sustainability of project be discovered, and will not be used (Panitch 2001, data in the online environment. 8). The survey results suggest that 15% of printed volumes, 27% of manuscripts, and 35% of the audio and video collections held by the 100 ARL respondents were unprocessed or uncataloged at that time. By comparison, a 2009 survey of a broader population of North American research libraries revealed that the situation has improved only marginally over the last decade, in spite of widespread focus on the hidden collections 20 Taking Stock and Making Hay: Archival Collections Assessment

Over the course of its three-year history, Practice with Purpose: Why the Cataloging Hidden Special Collections and Archives Program has awarded more Collections Assessment? than $11.9 million to a total of 46 projects The first step when considering a collections selected from approximately 300 proposals. assessment is a careful articulation of the In coordination with this program, CLIR reason—or reasons—for which it is to be maintains a web-accessible registry of hidden undertaken. Because even a small survey project collections, based upon information supplied is very likely to be a complex undertaking, and by applicants and others (CLIR 2011). because resource allocators are more likely to The National Historical Publications and support an effort that prescribes one or more Records Commission (NHPRC) has a long concrete outcomes, it is important to design history of providing funds for “fundamental the project in such a way that its objectives are archival activities” in the form of basic clear, its audience is apparent, and its benefits processing projects that “reveal collections are maximized. Depending upon its intended that researchers cannot easily discover” purpose and the resources allocated to it, a (NARA 2011). Institutions are required to collections assessment can range from a one- time-only of some or all holdings to a • create and share collection-level information; comprehensive, ongoing, data-gathering activity. • develop or implement appraisal, Most survey projects are undertaken for processing, and other techniques that one or more of the four purposes described will eliminate existing backlogs and/ below. Although none of these precludes or prevent future backlogs; and another, it is difficult to put equal emphasis • promote the use of processed collections. on all of them. Early in the project, therefore, The National Endowment for the Humanities’ it is essential to decide which goals or (NEH) Humanities Collections and Reference outcomes are considered primary, which Resources program (NEH 2011) supports efforts are considered secondary, and which will that “provide an essential foundation for not be addressed. The assessment projects scholarship, education, and public programming described later in this report have gathered in the humanities” by funding projects that information with the goal of accomplishing at address one or more of the following activities: least one of the four aims described below. • arranging and describing archival and collections; Expose Hidden Collections • cataloging collections of printed works, Many institutions have undertaken a collections photographs, recorded sound, moving assessment for the primary purpose of preparing images, art, and material culture; and sharing consistent, comparable, summary descriptions of some or all of the collections • providing conservation in their care. If this is indeed the primary treatment for collections; goal, the assessment activity may consist • digitizing collections; primarily of assembling, normalizing, and/or • preserving and improving access augmenting existing descriptive information to born-digital sources; and at the collection level; indicating whether or not the collection is available for research; • developing databases, virtual and making this information available— collections, or other electronic resources preferably online. More often than not, however, to codify information on a subject creating uniform collection-level descriptions or to provide integrated access to necessitates the gathering of information that selected humanities materials. can only be obtained by physically inspecting some portion of the collections and collecting information about those collections with a systematic, well-documented approach.

21 Making Archival and Special Collections More Accessible

Some of the institutions that have undertaken of the containers in which it is housed to a more collections assessments of this type have done detailed, systematic evaluation that provides so with the explicit intention of exchanging the institution with a better understanding of collection-level information with other the prevalence of specific conservation issues institutions and/or depositing collection-level as well as unmet preservation challenges. descriptions in a consortially—or regionally— Libraries and archives have a long history of using managed database (PACSCL 2009). a variety of well-established, well-documented methods to capture essential information about Establish Processing Priorities the current state and the ongoing needs of the collections in their care. Preservation surveys Especially for institutions with large backlogs of focus primarily on diagnosing large-scale and/ un- and under-processed collections, a collections or pervasive problems at the collection level assessment serves as a very useful tool for and assessing the overall storage and housing planning, informing, and guiding priorities for environment, usually to make the case for facilities collections processing. With this purpose as its improvements that will slow or prevent future primary goal, the collections assessment becomes damage. Conservation surveys tend to highlight a more complicated undertaking, as it requires the scope and the distribution of problems that collecting information and making judgments plague particular media (such as acetate film, about various aspects of the collection, only brittle paper, and deteriorating magnetic tape) some of which may already be known or are easily and support the allocation of limited resources determined. A collections assessment aimed for treatment. Increasingly, however, the at establishing processing priorities includes consideration of preservation challenges and/ but goes well beyond the gathering of basic or conservation issues is but one component information about the size, scope, and contents of of a larger balancing act, the overarching goal the collection. It typically requires that surveyors of which is to make collections accessible. assess the condition of the collection material as well as the containers in which it is housed; Because collections that cannot be handled determine the ease with which material in the physically without causing additional damage collection can be located; and evaluate the ease cannot be used, information about physical with which the collection can be discovered, condition is typically used to help answer identified as relevant, and used, based upon the basic questions such as “How is use of this existence and the accessibility of catalog records, collection hampered or limited?” and “Does finding aids, and other collection surrogates. the degree of damage or deterioration, or the value of the collection, justify reproduction or A collections assessment intended to inform the treatment?” For many institutions, however, assignment of priorities for processing should laying the groundwork for the establishment also include for each collection some kind of of preservation and/or conservation priorities estimation of its research value for present is the primary goal of the collections and future users. Techniques for determining assessment, warranting greater emphasis on research value are described in the “Collecting the comprehensive capture and systematic Qualitative Information” section of this report. tracking of essential information about condition (see, for example, 2011). Assess Condition Even if establishing preservation and/or Manage Collections conservation priorities isn’t the primary goal Much of the information gathered during a of a collections assessment, it is difficult to collections assessment can be used almost resist the opportunity to capture information immediately to address a number of collection about physical condition when a collections management issues including optimizing storage assessment is underway. This appears to be efficiencies, identifying strengths and gaps in the case for all types of institutions and across collecting areas, and validating de-accessioning all categories of collections. The information decisions. A comprehensive inventory is the typically gathered ranges from a basic assessment foundation of effective collection management, of the overall condition of collection material and and, when coupled with the value judgments that

22 Taking Stock and Making Hay: Archival Collections Assessment usually accompany a collections assessment, Will all collections be surveyed? Un- or under- provides a powerful tool for repositories with processed collections only? Or will other burgeoning backlogs of un- and under-processed criteria determine the scope of the collections collections, significant “information gaps” assessment? With the exception of those surveys regarding the contents of collections, and/or that have been undertaken primarily to amass pressing space concerns. Without exception, information about conservation issues, the scope those institutions that have undertaken of a collections assessment is typically limited collections assessment for any of the first three to un- and under-processed collections only. primary purposes described above have reaped Often there are good reasons to limit the scope inevitable secondary benefits in the form of better of an assessment undertaking to collections informed, more active collection management. consisting of or containing certain types of material or special formats (such as artworks, audio-visual material, photographs, realia, etc.). Ready, Set, Go! Conducting Increasingly, however, institutions are using collections assessment for purposes that require the Assessment a broader scope. Examples of purposes that Several important activities must be necessitate surveying all collections include accomplished before the survey team can get discerning changes to collection development to work. These include defining the scope of policies and populating a collection management the project; determining the methodology database (see, for example, UCB 2011). and the resources that will be employed; and documenting the policies and procedures that will govern the assessment. Collecting Information Collections assessment is essentially an information-gathering activity. It is centered on Scope the systematic collection of quantitative and Guided largely by the purpose—or purposes— qualitative data about various characteristics of the collections assessment, scope is a of collections, including extent and contents, fundamental consideration that must be condition, accessibility (physical and intellectual), determined at the start of the project and and research value. Its immediate result is an carefully managed throughout. Other factors array of data that makes possible the provision that should be taken into consideration when of adequate, consistent, collection-level determining the scope of the assessment descriptions; affords a better understanding include the availability of resources (human of unmet preservation needs; and informs and financial), time, and physical space. For important decisions regarding collection many institutions, some or all of these may management, processing priorities, and be limited, and the scope of the assessment selection and other activities associated undertaking should reflect that reality. with digitization and exhibit preparation.

23 Making Archival and Special Collections More Accessible

Methodology Collecting Quantitative Information The collections that have been identified for In collections assessment, quantitative methods assessment are likely to vary considerably in many are used to collect basic information about respects, including size, complexity, condition, the extent of each collection and the types and type of material. For each, however, the basic of materials of which it consists. Collecting approach is the same: open the boxes and look quantitative information should be relatively easy at the stuff. In keeping with the stated objectives and require little or no judgment. “How many governing the assessment, surveyors will do of what?” is the basic question; because it can some or all of the following for each collection: be asked—and answered—in a number of ways, however, it is important to consider the following: • count and assess the condition of the containers in which collection • Will every box be opened, or is material is housed; some form of sampling OK? • identify and assess the condition of the • How will extent be measured (items, material(s) of which the collection consists; boxes, linear feet, shelf feet)? • evaluate its arrangement in terms of the • How will content (in terms of types of ease with which material can be located; materials, special formats, etc.) be identified and categorized (checklists, guidelines, etc.)? • determine the existence and the accessibility of catalog records, finding aids, and Anticipating with good planning and other collection surrogates; and addressing with good documentation, these and similar questions are essential. • assess its research value. Clear instructions—including definitions, Collecting Qualitative Information illustrations and examples—for all of the above are essential. The survey tool, along with Qualitative methods are used to collect accompanying forms, checklists, etc., should information about the condition, physical be thoroughly tested before actual surveying accessibility (arrangement), intellectual begins. Although survey data may be recorded accessibility (description), and research on paper worksheets, it is typically stored value of the collection. Collecting qualitative in a relational database, such as FileMaker information usually requires making some kind Pro or Microsoft Access, where it can be of judgment in order to assign a rating (value) accessed and manipulated as necessary. along a numeric or descriptive continuum (scale). In a numeric continuum, 1 is usually the Staffing lowest or worst rating and 5 is the highest or best. In a descriptive continuum, values range Who will do the surveying? What do they need from, for example, “poor” to “excellent” or from to know? Although appropriately staffing the “negligible” or “none” to “significant” or “very assessment is an important consideration, high.” Ratings can (and should) be defined and and in some cases the availability of human documented in such a way that they can be resources may have the effect of defining or assigned systematically and consistently. Other limiting the scope of the assessment, successful possibilities for measures include estimating assessment projects have been accomplished the amount of collection material that meets with a variety of staffing models, ranging from a particular criterion, for example, “What those that employ experienced archivists, percentage of the collection needs new housing?” , and conservators (experts/professionals) to those that rely on individuals who have Condition some knowledge but no experience with the Assessing the physical condition of collection collections (generalists) to those that draw material and the quality of the housing in on a large body of students, volunteers, and/ which it is contained is often one of the most or others who have neither knowledge of nor important—and most difficult—activities in experience with collections (novices). In all cases, collections assessment. This is especially the case adequate training and good documentation for collections that contain or consist primarily of are key factors to a successful undertaking. material in formats other than paper. It is helpful 24 Taking Stock and Making Hay: Archival Collections Assessment to keep in mind that a collections assessment a collection is only truly accessible when is neither a preservation planning survey nor a researcher can find information about it a collection condition survey. As such, effort is online. The best rating, therefore, is reserved not usually dedicated to noting the condition for those collections that are described online; of particular items or to identifying groups of the worst is assigned to those that are not materials of particular concern, although these described at all or are described only in an may be called out in some way. Assessing physical accession record, donor/control file, or other condition and housing quality as one component document that is inaccessible to researchers. of a larger, more general collections survey is aimed at providing a better understanding of the Research Value overall condition of collection material, the overall Assessing research value is probably the quality of the boxes, folders, and other containers most troublesome component of collections in which it is housed, and the degree to which assessment for a number of reasons, most of one or both of these might or will hinder its use. which can be attributed to the difficulty—real or perceived—attendant in defining and measuring Arrangement “research value” in the first place. Assuming, Also potentially hindering the use of a collection however, that it can be defined and measured, a is its physical arrangement, which is one of the thoughtful assessment of research value usually reasons why collections assessment typically provides compelling information that can be includes an evaluation of the ease with which used to inform important decisions regarding material in the collection can be located. That collection management, processing priorities, evaluation usually takes into account both the and selection and other activities associated size and the complexity of the collection, does with digitization and exhibit preparation. not assume that arrangement to the item level In the collections assessment context, the term is necessary or desirable, and is focused on “research value” usually refers to the value of rating the collection in terms of how successfully the collection in terms of the extent to which it can be used for research. A small, relatively it includes relatively rare, extensive, and/or homogeneous collection in rough order, for detailed information about a topic that has example, is generally more physically accessible received considerable prior attention, is gaining than a large, heterogeneous collection in rough , and/or has apparent potential to attract order, and the ratings should reflect that. significant interest. It is frequently expressed as Description a composite of two measures: documentation interest and documentation quality. Before a collection can be used, however, it must be discovered and identified as relevant. The documentation interest rating provides Both of these depend upon the existence and an indication of the value of the collection in the accessibility of catalog records, finding terms of its topical significance, with values aids, and other collection surrogates. Rating ranging from 1 (negligible or none) to 5 (very the “intellectual accessibility” of a collection, high). Similarly, the documentation quality then, typically requires determining if and how rating provides an indication of the value of the well the collection is described (in an accession collection in terms of its topical richness, again report, catalog record, finding aid, etc.) and with values ranging from 1 (slight) to 5 (very evaluating the accessibility—especially the online rich). Research value can then be expressed as availability—of any existing descriptions. While a the sum of the documentation interest rating catalog record and/or a simple inventory might and the documentation quality rating. provide adequate access to a small, relatively Here it must be pointed out that many collections straightforward collection, neither is sufficient have values in addition to, or even other than, for a large or complex collection. Finding aids research value, and that these can—and typically provide the best intellectual access should—be measured if the overall purpose of to archival collections, especially large and/or the assessment warrants such an evaluation. complex collections, and the ratings that are Examples of these “other” values include assigned during this component of a collections intrinsic value and local or institutional value. assessment are governed by that assumption. The ratings also reflect the expectation that 25 Making Archival and Special Collections More Accessible

Collecting “Other” Information Putting it All Together In addition to collecting the above-described Of course, the reason for collecting assessment information about the collection, many data is to put it to use. Here are some examples institutions also collect information about of collection assessment in action. the assessment process itself, including who The following example (figures 1 and 2) from the conducted the assessment, when it was University of Michigan Special Collections Library accomplished, how long it took and if any shows that while the collection rates fairly high activity (such as reboxing) was undertaken. in terms of research value, it is intellectually inaccessible and physically difficult to use.

Figure 1. University of Michigan Special Collections Library collection sample

Figure 2. University of Michigan Special Collections Library ratings sample

26 Taking Stock and Making Hay: Archival Collections Assessment

The Smithsonian Archives uses collections assessment to measure and demonstrate change over time in their preservation module. Represented in figure 4, the initial assessment of this collection in 2001 (Accession 000182 United States Civil Service Commission) shows an Overall Priority score of 2.

Figure 3. Smithsonian Archives preservation model: initial assessment of United States Civil Service Commission 2001 collection (Accession 000182)

27 Making Archival and Special Collections More Accessible

In 2003, the collection was re-assessed after preservation actions were taken (shown in figure 4). Some items were discovered to be rolled while reboxing the collection and a score of 2 was given to the Difficult Formats/Sizes category. While the overall Priority Score for this collection did not change at this time, the Holdings Maintenance Score went from a 6 to a 2.

Figure 4. Smithsonian Archives preservation module: reassessment of United States Civil Service Commission 2001 collection (Accession 000182)

28 Taking Stock and Making Hay: Archival Collections Assessment

Finally, the entire collection was assessed before moving the collection offsite in 2006. This generated a new overall Priority Score of 5 (shown in figure 5).

Figure 5. Smithsonian Archives preservation module: final assessment (pre-offsite move) of United States Civil Service Commission 2001 collection (Accession 000182)

29 Making Archival and Special Collections More Accessible

The Logjam Project (Northwest Archives Council, UK) provided an “audit toolkit” as a data collection tool (shown in figure 6). The toolkit provides a calculation that generates an estimate of Cataloging Resources for collections, taking into account such factors as the Extent of the collection, main Covering Dates, Level of Cataloging necessary and the potential Complexity of cataloging the collection*. Each data field is weighted and a Resources Score is generated, producing a cataloging estimate for each collection, with times given for both professionals and paraprofessionals. Calculations are based on regional norms, but could be adjusted for other circumstances.

Figure 6. Logjam audit kit sample from Northwest Archives Council, UK

*Logjam gives definitions for values for “level of cataloguing” (High—Uncatalogued; Medium—Box-listed; Low—Listed to series level; and “cataloguing complexity” (Very Complex, Complex, Moderate, Moderate Straightforward, Straightforward).

Assigning a research value rating as a component What’s Missing from of archival collections assessment is one way This Picture? to estimate potential scholarly significance; another might be mining use and other data By providing both the opportunity and a process to determine how heavily used a particular for documenting a wide range of characteristics collection, or group of collections, is, especially about the collections in our care, archival in relationship to other collections. Although collections assessment can be used to address a recording “amount and type of use” is not variety of important needs, including collection typically integrated into collections assessment management issues and processing priority- activities, and would in fact require data external setting. Existing practices, however, do not fully to the survey process per se, it should feature support other equally-pressing concerns. more prominently in our user-centric world.

Researcher Needs Collection Development Policies Much of the focus of archival collections The 2009 survey of special collections and assessment is oriented to the needs of the archives in North American academic and collections themselves. Which require rehousing? research libraries paints a picture that is both Need basic conservation? Lack adequate encouraging and sobering. Among the key description? The ever-increasing emphasis in findings described in the report are the following: libraries and archives on meeting the needs of researchers—for whom we have collections • monetary resources are shrinking; in the first place—will likely result in less • collections, and user demand for support for assessment activities that do not access to them, are growing; and include the identification of collections that are expected to be of high research interest. • space for collections is inadequate. This “trifecta” of sorts serves to remind us of the importance of acquiring and devoting resources to the needs of those collections that are most valuable from a research perspective and that fit best within existing collection strengths. The fact that few institutions are likely to 30 Taking Stock and Making Hay: Archival Collections Assessment secure additional storage capacity, and even fewer are likely to stop collecting, underscores the importance of collecting policies. Only half of the respondents in the 1998 survey of special collections in ARL libraries indicated that they have formal collection development policies. By revealing existing collection strengths, collections assessment can serve as a powerful motivator for those institutions that need to develop and/or refine meaningful collecting policies. Collections assessment data can also be used, when necessary, to make a case for deaccessioning “out of scope” and “not a good fit” collections.

Digitization Readiness In a world that is increasingly shaped by the view that “if it isn’t online it doesn’t exist,” digitization of special collections material is—or should be—at or near the top of our priority list. Although some of the data gathered during the course of a “typical” collections assessment contributes significant value to the selection for digitization process, other important data is not usually collected. By anticipating the need to answer questions about , access and/or use restrictions, and the extent to which a particular collection (or related collections) has already been digitized, archival collections assessment can play a critical role in helping us move forward in this important arena.

Conclusion The combination of almost limitless collecting opportunities and increasingly limited resources with which to get the job done requires that we identify, articulate and focus our attention on the priorities that are most central to our mission. Whether undertaken as a one-time, for-one-purpose-only project or integrated into an overall approach to managing collections, archival collections assessment can help us set those priorities by taking much of the guesswork out of the picture.

31 Making Archival and Special Collections More Accessible

Appendix A: Project Descriptions

The Black Metropolis Worksheet (in Excel) designed for use in a comprehensive collection risk assessment. Research Consortium

Consortial Survey Initiative of African American History Museum Materials (January 2009-December 2011) http://lucian.uchicago.edu/blogs/bmrcsurvey/ Manuscripts Cataloging, Survey, and Processing Project (October 2009-March 2010) With funding provided by The Andrew W. Mellon Foundation, the Chicago-based Black Metropolis For more information, e-mail Peter Research Consortium (BMRC) is undertaking a Alter ([email protected]) comprehensive survey of collections of materials With funding from the National Historical that document African American and African Publications and Records Commission (NHPRC), diasporic culture, history, and politics held by its the Chicago History Museum conducted a 14 member institutions and by 20 community- cataloging and assessment survey of the based African American organizations and Museum’s archival and manuscripts holdings. creators. The goals of the survey are several Informed by projects at the Historical Society and include making possible the creation of of Pennsylvania and the Area preliminary descriptions of collections that Consortium of Special Collections Libraries are inaccessible to researchers; informing (PACSCL), the survey methodology was modified the prioritization of preservation and access to support the Museum’s MPLP-based, tiered needs; and enabling collaboration, building processing approach. The purpose of the survey partnerships, and sharing “best practices” was to (1) create and/or verify and enrich catalog between and among initiative participants. records for all collections of half a linear foot or The project website includes links to survey larger; (2) assess holdings to determine each documentation, status reports, and the Second collection’s ideal minimal processing level (i.e., Space Initiative, which facilitates access to collection, series, sub-series, or folder); (3) relevant research material held outside the identify un- and under-processed collections (4); library, museum, and archival communities. prioritize collections for processing and (5) flag “” material (unaccessioned Canadian Museum of Nature collections and/or collections with inadequate accession documentation). Project staff verified Assessing and Managing Risks and enhanced more than 1,000 catalog records to Your Collections describing the Museum’s manuscript holdings, created approximately 30 new catalog records, http://nature.ca/en/research-collections/ and generated a non-public database to our-scientific-services/assessing- guide the planning and management of future managing-risks-your-collections preservation and processing activity. More than 100 collections (totaling nearly 1,300 linear feet) The Canadian Museum of Nature regularly were processed to the series-level in a later offers workshops on identifying, ranking, and phase of the project, and approximately 300 mitigating risks to collections of cultural property. “found in collection” problems were resolved Based upon the Cultural Property Risk Analysis through legal and/or administrative measures. Model developed by the Canadian Museum of Nature, the workshop provides participants with a methodological approach to identifying types of risk, calculating magnitudes of risk, determining methods for controlling risks, and evaluating mitigation strategies. Participants receive a manual and a Risk Assessment

32 Taking Stock and Making Hay: Archival Collections Assessment

Columbia University Historical Society of Pennsylvania

Mellon Special Collections Mellon Collections Preservation and Backlog Materials Survey (2003-2004) Processing Planning Project (2000-2002) http://www.columbia.edu/cu/lweb/ http://www2.hsp.org/collections/ services/preservation/surveyTools.html manuscripts/Mellon/about.html Between October 2003 and July 2004, staff This comprehensive survey appears to be at Columbia University Libraries surveyed the first in a series of Mellon-funded projects unprocessed collections held in the Rare aimed at collecting qualitative and quantitative and Manuscript Library, the Avery Architectural data about unprocessed special collections and Fine Art Library, and the C.V. Starr East Asian material. The model developed in this project Library. In total, 1,588 survey hours were spent includes measures of the following for each entering data on 569 collections and accounting collection surveyed: physical condition, quality for 26,299 units stretching 15,867 linear feet. of housing, physical access (arrangement), These collections are composed of 8,703 feet of intellectual access (description), and research loose paper; 87,948 bound volumes of all types; value (interest and quality of documentation). 100,903 architectural drawings; 14,218 graphic Between 2000 and 2002, project staff surveyed works; 158,478 photographic materials; 136,457 approximately 5,000 collections, including 3,000 negatives, slides, motion pictures and microfilm; manuscript collections; 300,000 maps, prints, 1,288 phonographs; 6,559 audiotapes, videotapes drawings, broadsides, and photographs; and and computer media; 277 optical media items, approximately 11,000 art objects and artifacts. and nearly 3,400 pieces of realia and memorabilia. The project website includes a guide to the survey North West Regional instrument/database and a description of the ratings, both of which are modeled on, but vary Council (UK) from those developed at the Historical Society of Pennsylvania. Staff in the Rare Book and Logjam: An Audit of Uncatalogued Manuscript Library are using the database to track Collections in the North West accessions and as a source for box lists and other forms of preliminary and intermediate access http://www.northwestcultureobservatory. tools; Preservation Department staff rely on it for co.uk/ preservation and conservation planning and for [You must create a free account. Once you establishing departmental goals and priorities. are logged in, search for “Logjam” in the databank to download documentation.] Taking the form of a detailed audit, the Logjam project was designed to “scope the size and type of uncatalogued collections held in 30 of the region’s principle archive-holding institutions.” The work was undertaken by the North West Museums, Libraries and Archives Council (NWMLAC) on behalf of North West Regional Archive Council (NWRAC). The project represents one component of a strategy aimed at improving and expanding access to the region’s archives by making finding aids and collections more widely available and by developing a collaborative approach to cataloging backlogs. Specific goals of the project include (1) producing a detailed picture of the uncataloged archival collections held in each repository and in the region as a whole; (2) describing the resources required to catalog these

33 Making Archival and Special Collections More Accessible collections; (3) assigning priorities for cataloging Angels Project connects conservators with a these collections and (4) identifying priorities for collection that needs care. The project described future collaborative projects within the region. in this report served as a pilot to develop and demonstrate “ideal” procedures for the processing, rehousing, and reformatting of an Philadelphia Area Consortium important collection of scientific illustrations. of Special Collections Libraries (PACSCL) Smithsonian Institution Archives

Consortial Survey Initiative (2006-2008) Preservation Assessment Component http://www.pacsclsurvey.org/ of Collection Management System The Philadelphia Area Consortium of Special For more information, e-mail Sarah Collections Libraries (PACSCL) Consortial Stauderman ([email protected]) Survey Initiative is a 30-month project funded The Preservation Assessment Component by The Andrew W. Mellon Foundation to provides a mechanism for the capture and assess unprocessed, underprocessed, and tracking of essential information about the underdescribed archival collections in a range condition of collections. Seven questions guide of physical formats held in 22 Philadelphia area the assessment; answers (provided on a scale institutions. Modeled on the Historical Society of from 0 to 3) are used to automatically calculate Pennsylvania project, the purpose of the survey and assign preservation priority. Assesses the is to collect data that can be used to (1) inform percentage of the collection that needs housing; planning for, and prioritization of, collections has inappropriate housing material (e.g., acidic work within individual institutions and across folders, envelopes); is poorly positioned (e.g., the consortium and (2) improve intellectual messy, overstuffed); has format problems (e.g., access to unprocessed and underprocessed crushed, folded, rolled); has damaging and/ collections by making collection-level records or inappropriate attachments (e.g., staples, available to the public. As of the end of October, paper clips, etc.); has physical damage (from 2,100 collections totaling over 19,400 linear dirt, adhesive, water, etc.); and has unstable feet in 22 institutions have been surveyed. materials (e.g., newspaper, thermo fax paper, Survey data is recorded in a shared, publicly- color photographs, etc.). Also provides accessible database developed specifically mechanisms for recording actions taken for the project. Because it includes fields during the assessment, including rehousing that allow institutions to maintain internally and digitization, and for alerting preservation significant data, such as location and staff to immediate and/or long-term needs that information, the database can be used as a cannot be addressed during the assessment. basic accessions or collection management system by individual institutions. The website includes links to project documentation University of California, including a survey checklist, a description of Berkeley, The Bancroft Library the ratings, and a guide to the database. Manuscripts Survey Project Smithsonian Institution, National (February 2008-January 2011) Museum of Natural History http://blogs.lib.berkeley.edu/bancsurvey.php With funds provided by The Andrew W. Mellon Angels Project (1996) Foundation and the Rosalinde and Arthur Gilbert Foundation, the Bancroft Library is undertaking http://cool.conservation-us.org/coolaic/ a comprehensive survey of all manuscript sg/bpg/annual/v15/bp15-18.html holdings processed before 1996, including a In conjunction with the annual meeting of backlog of some 595 collections representing the American Institute for Conservation, an 25,000 linear feet of archival and manuscript 34 Taking Stock and Making Hay: Archival Collections Assessment material that is currently unavailable for University of Michigan research. Project staff, working over a three-year period, will apply standard archival appraisal methodologies to each collection in order Unprocessed Collections Survey Project (2009) to determine its scope and content, identify For more information, e-mail Martha preservation needs, make recommendations Conway ([email protected]) regarding arrangement and description, and estimate the resources required to make it fully This project engaged masters-degree students accessible to researchers. The survey will yield at the School of Information in two consecutive updated, accurate, and detailed information projects surveying un- and under-processed that will be used to establish processing goals, collections held by the Special Collections develop funding priorities, and facilitate collection Library. Working in teams of two and three, using management decisions, including those involving an assessment methodology derived from the the de-accessioning of out-of-scope materials. Historical Society of Pennsylvania project, 55 students collected quantitative and qualitative information on a total of 40 unprocessed collections of archival and manuscript material. Their findings, and the reports documenting their effort and their observations, have been used to populate a web-accessible database that the Special Collections Library will employ to create and make available adequate and uniform collection-level descriptions; understand more fully the prevalence of unmet preservation challenges; inform collection management decisions; and establish and guide processing priorities. Project documentation includes a field-by-field description of the database tables and an illustrated procedure manual.

35 Making Archival and Special Collections More Accessible

University of Virginia condition of processed manuscript and photograph collections held by Manuscripts, Archives, and Special Collections. Andrew W. Mellon Special Collections Assessment Project (2002-2004) Staff surveyed approximately 4,400 linear feet of manuscripts and 120 collections containing https://www.lib.virginia.edu/ more than 500,000 photographic images. The small/collections/mellon/ project website includes links to survey forms, sample database records, and a photo gallery. Modeled on the Historical Society of Pennsylvania project, this survey of the archival and manuscript holdings in the Special Collections Library WGBH Media Library and Archives resulted in data that has been used to determine cataloging and processing priorities and to generate time and cost estimates for the work Assessment for Scholarly Use associated with collections that require additional http://openvault.wgbh.org/pdf/ processing. Project staff collected several types WGBHMLAAssessment.pdf of use data to measure current interest in the holdings of the Special Collections Library, The WGBH Media Library and Archives (MLA) developed a methodology to identify current Assessment for Scholarly Use project, funded and future research trends that those collections by The Andrew W. Mellon Foundation, was might support, and evaluated the ease with which designed to achieve two goals: to determine the staff can locate and serve collection materials educational value of WGBH’s extensive archival to patrons and the ability for patrons to identify collection for higher education research and relevant materials in those collections. The instruction, and to accomplish this by designing survey procedure manual and the data collection an assessment instrument for surveying audio- form are available at the project website. visual collections that could be shared with other institutions. At the start of the project, the MLA housed approximately 29,000 programs Washington State with 570,000 related production elements and University Libraries documents, numbers that suggest both the potential worth of this collection to the academic community and the complexity of evaluating Comprehensive Preservation Survey of its educational value. The study approached Manuscript and Historical Photograph this challenge by (1) creating a framework Collections (2004-2005) and tool for collecting information about the archived programs (2) assembling a detailed http://www.wsulibs.wsu.edu/holland/ composite portrait of the archival collection and masc/preservationsurvey.html (3) modeling potential approaches to analyzing With a grant from the Washington Preservation and employing the data compiled through this Initiative, a LSTA-funded program administered work. The extensive project report includes by the Washington State Library, the Washington the assessment tool and recommendations State University Libraries assessed the physical regarding its use by other institutions.

36 Taking Stock and Making Hay: Archival Collections Assessment

Appendix B: Procedure Manual (University of Michigan) Available online: http://www.oclc.org/research/activities/backlogtools/michiganmanual.pdf

Appendix C: Ratings Descriptions (Columbia University) Available online: http://www.oclc.org/research/activities/backlogtools/columbiaratings.pdf

37 Making Archival and Special Collections More Accessible

References

ARL (Association of Research Libraries). 2008. PACSCL (Philadelphia Area Consortium of Special Collections Task Force Final Status Special Collections Libraries). 2009. “PACSCL Report, 2006. Last modified 4 April. http://www. Consortial Survey Initiative.” Last modified arl.org/rtl/speccoll/spcolltf/status0706.shtml 15 March. http://www.pacsclsurvey.org. ---. 2009. “Transforming Research Libraries. Panitch, Judith M. 2001. “Special Collections The unique Role of Special Collections. in ARL Libraries: Results of the 1998 Survey Exposing Hidden Collections: 2003 Conference Sponsored by the ARL Research Collections Summary. Hidden Collections Conferees Committee.” Washington, DC: Association Call for Local and Collective Actions” Last of Research Libraries. http://www.arl. modified 6 April. http://www.arl.org/rtl/ org/bm~doc/spec_colls_in_arl.pdf speccoll/EHC_conference_summary.shtml. Pritchard, Sarah M. 2009. “Special CLIR (Council on Library and Information Collections Surge to the Fore.” Libraries Resources). 2011. “Cataloging Hidden Special and the Academy. 9(2): 177—180. Collections and Archives.” http://www. Schreyer, Alice. 2007. “University of Chicago clir.org/hiddencollections/index.html. Explores Library-Faculty Partnerships Columbia University. 2011. “Columbia University in Uncovering Hidden Collections.” ARL: Libraries Preservation and Digital Conversion A Bimonthly Report 251 (April): 4. http:// Division. Special Collections Materials Survey www.arl.org/bm~doc/arlbr251.pdf Instrument.” http://library.columbia.edu/ Steele, Victoria. 2008. “Exposing Hidden services/preservation/survey_tools.html Collections: The UCLA Experience.” C&RL Dooley, Jackie M., and Katherine Luce. 2010. News 69 (6). http://www.ala.org/ala/mgrps/ “Taking Our Pulse: The OCLC Research Survey divs/acrl/publications/crlnews/2008/ of Special Collections and Archives.” Dublin, jun/hiddencollections.cfm. Ohio: OCLC Research. http://www.oclc.org/ Tabb, Winston. 2004. “‘Wherefore Are These research/publications/library/2010/2010-11.pdf Things Hid?’ A Report of a Survey Undertaken Greene, Mark A. and Dennis Meissner. 2005. “More by the ARL Special Collections Task Force” Product, Less Process: Revamping Traditional RBM: A Journal of Rare Books, Manuscripts, Archival Processing.” The American . and Cultural Heritage 5 (2): 123-126. http:// 68 (2): 208—263. http://archivists.metapress. rbm.acrl.org/content/5/2/123.full.pdf. com/content/c741823776k65863/fulltext.pdf UCB (University of California, Berkeley). 2011. “The Hewitt, Joe A. and Judith M. Panitch. 2003. Bancroft Library Manuscripts Survey Project” “The ARL Special Collections Initiative.” Library Last modified 18 February. http://bancroft. Trends 52 (1):157—171. http://www.ideals. berkeley.edu/info/mss_survey/index.html. .edu/bitstream/handle/2142/8497/ librarytrendsv52i1p_opt.pdf?sequence=1. NARA (National Archives and Records Administration). 2011. “National Historical Publications and Records Commission (NHPRC).” FY 2011 Grant Announcement: (initial). Archives—Basic Projects. http://www.archives. gov/nhprc/announcement/basic.html. NEH (National Endowment for the Humanities). 2011. “Humanities Collections and Reference Resources”. Posted 6 April. http://www. neh.gov/grants/guidelines/HCRR.html.

38 Making Archival and Special Collections More Accessible 5 Over, Under, Around, and Through: Getting Around Barriers to EAD Implementation

Michele Combs

Mark A. Matienzo

Merrilee Proffitt OCLC Research

Lisa Spiro Rice University

This paper was originally published in February 2010 by OCLC Research at http://www.oclc.org/content/dam/research/publications/library/2010/2010-04.pdf.

39 Making Archival and Special Collections More Accessible

Introduction

This report frames some of the obstacles documented. For every roadblock, as Sesame that archivists have experienced adopting Street’s Grover says, there is a way “over, under, Encoded Archival Description (EAD). It also around, and through.”6 This paper presents useful suggests pathways to help you get out of the tools—informational, persuasive, or technical— ruts, around the roadblocks, and on the road for overcoming barriers you may encounter in to success. This report is addressed to those your journey towards EAD implementation. who have a basic understanding of standard Section I of this report addresses political and archival descriptive structures and modest logistical issues. These include gaining buy-in acquaintance with EAD. Our objective is to from institutional decision makers, overcoming help you communicate EAD’s value as a key the urge to achieve perfection, finding ways to element of successful archival information maximize scarce resources, and getting over systems and overcome potential barriers to its the initial humps of dealing with a relatively implementation. This paper does not contain complicated standard and what can be an EAD primer, or cover the basics of document perceived as overwhelming logistical issues. encoding. For those who are not familiar with EAD, we recommend the EAD Help Pages as an Section II navigates technical problems and excellent starting place for more information.1 solutions, such as thinking about lossless data streams in conversion and management, Archivists have been encoding finding aids using selecting software (and challenges around open EAD for over a decade. An impressive number of source software in particular), publishing, and institutions have implemented EAD, but many mitigating the complexity of the standard. have not. A 2008 survey revealed that nearly half of respondents (79 out of 168) had not yet Members of this working group (under the implemented EAD.2 A further analysis of the auspices of the RLG Partnership and OCLC characteristics of those who had not yet adopted Research) authored this report jointly. We all have EAD reveals that all types of institutions are had experience with EAD and have struggled represented, including archives affiliated with with the range of issues. Thus, the advice we large and small universities and those with a offer comes from practical experience. range of information technology (IT) services, This paper addresses a wide range of needs from no professional IT staff to those with access because of the assortment of issues. We hope to the services of a large IT department. that you will dip directly into the sections that Our professional literature articulates obstacles are most appropriate to your particular need. ranging from political to technical, and much in We present barriers as articulated in published between. Over the last ten years a growing body of literature. We then propose one or more solutions relevant articles detail barriers: Jill Tatem’s article that may work for you. Our goal is to show you “EAD: Obstacles to Implementation, Opportunities that implementing EAD is easier than you think. for Understanding”; James M. Roth’s “Serving up We hope these strategies will be helpful and will EAD: An Exploratory Study on the Deployment smooth the way to successful implementation. and Utilization of Encoded Archival Description Finding Aids”; and Elizabeth H. Dow’s “EAD and the Small Repository.”3 These early works were followed by Katherine M. Wisser’s EAD Tools Survey and Sonia Yaco’s article, “It’s Complicated: Barriers to EAD Implementation.”4 Political or logistical issues may keep you from getting going; technical issues may get you bogged down along the way. Against this backdrop of challenges, there are a growing number of tools that support EAD.5 Nevertheless, real and perceived barriers to EAD implementation still exist, all of them well 40 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Section I: Political and Logistical Issues

I’m preaching to the unconverted7 Archivists).8 EAD is global; EAD has been implemented by a wide variety of institutions, not only in the US and Canada, but also throughout Solution: Prepare effective arguments Europe, , and Asia. about EAD’s significance EAD plays well with others EAD has been mapped to and from other data The following arguments may help you encoding standards such as MARC and Dublin communicate that EAD is a good investment Core.9 Because EAD supports hierarchical of institutional funds and staff resources. description, you can map data from a relational We begin with a brief “elevator speech” database; many commonly-used EAD tools to introduce the nature and purpose of are, in fact, built on relational databases. EAD EAD, followed by more specific points. need not be the environment in which you produce, store and manage your description, The elevator speech—What is EAD, and but it works well as a global transfer syntax. why should my institution invest in it? EAD encoding facilitates aggregation EAD is an international standard for encoding It would be difficult, if not impossible, to create finding aids established to meet the needs of both effective subject gateways like the American end-users and archivists. EAD is represented in Institute of Physics’ Physics History Finding XML (Extensible Markup Language), a platform- Aids Web site, or regional collection gateways neutral data format that ensures data longevity such as the Online Archive of California, when migrated from one software environment without the consistency imposed by EAD. to another. EAD ensures the long-term viability of your data by encoding intellectual rather An abundance of tools support than only presentational data (HTML, for EAD implementation example, only accomplishes the latter). EAD Tools exist to facilitate every aspect of EAD can be produced from (or mapped to) a variety use, from encoding to publication. So many of formats, including relational databases, tools exist that we’ve included only a selection MARC, , HTML and others, which in Appendix II. An even wider variety of makes it an excellent format for porting data. tools are covered in Archival Management In addition researchers can have a more robust Software: A Report for the Council on Library interaction with EAD finding aids because and Information Resources (2009).10 EAD enables better searching and subsequent delivery from a single source document. EAD implementation is supported by significant …and more! opportunities for training and collaboration Pick and choose from among the Opportunities abound for formal and informal following ideas that will be the most EAD training, advice and consultation to support persuasive in your circumstances. the growing population of EAD implementers. Some examples include the EAD discussion EAD is an internationally-used list, courses offered by the Society of American encoding standard Archivists and , and workshops at local, regional and national conferences.11 EAD complies with data content standards Various state and regional consortia offer EAD such as ISAD-G (the General International training opportunities, tools, and guidelines. Standard Archival Description, developed by the International Council of Archives) and EAD is good for researchers DACS (Describing Archives: A Content Standard, developed by an international working group ...in a number of ways: under the auspices of the Society of American

41 Making Archival and Special Collections More Accessible

1) Researchers can discover collections in EAD gets you money more places through wider availability. Grant agencies and other funders look favorably EAD’s consistent coding and structure on and encourage EAD implementation as means it’s easy to submit your finding part of their granting process. For example, aids to multiple access points (to the the guidelines for the NEH Preservation and Online Archive of California, or to OCLC’s Access, Humanities Collections and Resources ArchiveGrid, or to a subject-based portal encourage the use of EAD.14 The NISO/IMLS A such as the one maintained by the Niels Framework of Guidance for Building Good Digital Bohr Library & Archives at the American Collections includes EAD as an appropriate Institute of Physics, for example) so they’re metadata scheme for archives.15 NHPRC more likely to be found by researchers. similarly endorses EAD in their guidelines.16 2) Inexperienced researchers can use finding aids more easily. Consistency of Knowledge gained mastering EAD content and presentation eases the use of is applicable in other contexts collection descriptions for inexperienced In learning EAD, you will also develop skills researchers. Finding aids that are exposed that extend beyond encoding finding aids by online are far more likely to be found by gaining a basic understanding of XML and XML inexperienced researchers—an audience tools. So much digital data—in the library and whose needs we must always bear in archival communities and beyond—is stored mind—than collection descriptions and/or exchanged in the form of XML. These 12 that are only available locally. As user skills for staff will allow them to work with other studies reveal better and more intuitive standards such as MARCXML, MODS, and METS. ways to present finding aid content, reformatting collection guides encoded EAD paves the path to the future in EAD is painless. If one presentation/ Although today’s researchers find collection display method proves problematic descriptions using keyword searching on or confusing for researchers, you search engines, the Web of the future will be no can change it with minimal time and place for unstructured data. The future is the effort and zero rekeying or editing. “semantic Web” or linked data. Implementing 3) Researchers can filter and refine searches. EAD will help to position your institution Some applications can utilize EAD’s for the future of internet applications. structured tags. This makes it possible to limit searches to scope and content Everybody’s doing it! notes or collection titles, for example. Recognition of EAD’s significance has become 4) Display and output can be tailored for increasingly widespread, both within the US research needs. One single EAD encoded and internationally, and its use has expanded file can provide multiple output versions accordingly. The EAD Help Pages include for multiple researcher needs (online a comprehensive list of various types of version, printer-friendly version, etc.). institutions that are currently successfully You can also easily create different implementing EAD.17 We encourage you to display options for different audiences. look for institutions with a profile similar to yours. Knowing that your peer institutions 5) Researchers can explore old data in are implementing may help you persuade new ways. EAD enables archives to offer those that hold the purse strings that EAD researchers new, interesting, powerful, and implementation is worth the resources it will take. productive visual explorations of collections. There are some great new tools under development. Examples include: Jeanne Let me just tidy this up first…18 Kramer-Smyth’s ArchivesZ, an “elastic list” prototype at Syracuse, and relationship Encode the data you have mapping tools such as NNDB Mapper.13 Solution: to provide minimum access.

42 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

As wisely observed, “the perfect is the approach may help your institution get a toe in the enemy of the good.”19 The desire to achieve door with a modest investment of time or money. perfection can sometimes get in the way of small improvements that iteratively help us to reach Join the club larger goals. In the case of providing better access to our collections, the urge to rewrite finding aids Numerous state-, regional- or subject-based (or reprocess collections and then rewrite finding consortia have pooled resources to benefit aids) is a huge barrier to providing interim access member institutions by lowering barriers to EAD to the collection descriptions as they are now. implementation. They generally offer a range of services such as best practice guidelines, You must make every effort to make existing stylesheets, templates and other tools, training, collection descriptions as accessible as possible, and hosting of data. They often apply for grant regardless of your intentions for them in the funding to convert finding aids or provide other future. Although technically these collections services at little or no cost. Some consortia do not represent a processing backlog since may be able to handle all of your needs, they do have descriptions, if those descriptions including encoding. Even if you don’t belong are not accessible, they present the same to a consortium, many of these organizations problems as unprocessed collections. They are make their tools and guidelines freely available hidden from all but those inside the institution. on their Web sites, so that others need not A survey conducted in 2003-2004 by Dennis reinvent the wheel. Many consortial projects are Meissner and Mark Greene as background grant-funded; the more contributors and users for the “More Product, Less Process” report they have, the more likely they are to continue found that backlogs are a key concern for the being funded. Contributing your EAD thus helps majority of donors, researchers, and resource not only you but many other institutions and allocators.20 SAA’s code of ethics reminds us patrons. A partial list of regional, national and that “Archivists strive to promote open and international consortia is included in Appendix I. equitable access to their services and the records in their care without discrimination or preferential treatment.”21 Likewise, the SAA/ Take the first step ALA Joint Statement on Access includes the If you have collection-level records in the MARC following statements: “A repository is committed format, consider creating basic EAD records to preserving manuscript and archival materials through export (easily done using a tool such as and to making them available for research as MARCEdit).24 You will then have a set of collection- soon as possible” and “As the accessibility of level EAD data, and some experience working material depends on knowing of its existence, with the standard. The resulting files are also it is the repositories responsibility to inform suitable to contribute to an EAD consortium. 22 researchers of the collections in its custody.” You may then decide to expand the minimal In an increasingly online world, making your records, or you may decide to live with the fact collection descriptions as accessible as possible that your EAD descriptions will not include to the widest possible audience is of paramount inventories. Something is better than nothing, importance. Access deferred is access denied. particularly from the user perspective.

Who will do the work, and when?23 Take an iterative approach There is no rule that EAD encoding must be done once and only once, nor that it cannot be done Solution: Find low-impact ways to until arrangement and description are complete. tackle EAD implementation. Implement EAD with a “More Product, Less Process”-like approach! A collection-level finding aid with minimal information can be produced at Your staff is already stretched to the limit, and the time of accession. Areas such as bioghist adding EAD implementation to an already bulging and scopecontent can be expanded later. If workload is rarely feasible. Taking a creative the collection comes with a simple box list, that can be included at the outset, to be replaced

43 Making Archival and Special Collections More Accessible later by a fuller inventory when processing Templates is complete. Some collection management Templates are EAD documents pre-populated tools will produce collection descriptions as a with text that never changes (repository name, byproduct of the arrangement and description address, etc.) and with boilerplate text guiding the process. EAD finding aids and online publishing encoder to fill in proper data. This removes much free us from the static paper finding aid, offering of the angst of choosing what tags to use and how instead an evolving document that changes to use them, making it more like filling out a form. and grows though the life of the collection. Templates are simple to create. Using commercial Use someone else’s time and talent XML authoring tools, you can create an EAD file with as much information as possible— Do you have access to library school students, or including both text and attributes—already other interns? Do you have an internship program? filled in, and use it as the basis for all new EAD Consider offering an internship (or a series of files. For example, in the controlaccess internships) that focus on researching options section include one each of the possible child and implementing EAD for your institution. elements (subject, persname, corpname, etc) with the appropriate @source and @ You don’t have the time, but encodinganalog attributes filled in. Or in the you do have the money bioghist section, include the phrase “MARC 545: Insert brief bio or company history.” Some Maybe you don’t have staff time, but you can example of templates are included in Appendix II. make a one-time or ongoing investment of funds. If wholly or partially outsourcing is The EAD Schema of interest, there are more details below. Using the schema rather than the DTD version of EAD during authoring allows you to enforce 25 It’s so complicated! various content limitations, such as correct formatting of @normal attributes for date elements, thus reducing the chance for errors. Solution: There are many options While leveraging the full power of the EAD schema that make EAD simple may require a more substantial investment of time, schema-based validation can be used in combination with templates to ensure strict EAD can intimidate even tech-savvy internal compliance. An example of institution- staff, given the number of tags and the wide best practice guidelines implemented seemingly endless variety of ways they can using the XML Schema version of EAD is that be implemented, but not all elements are of Yale University’s Finding Aid Coordinating required. Use collection-level descriptions Committee. Yale finding aids validate against the and minimum-level description elements—as W3C Schema version of EAD, and compliance given by DACS—to simplify EAD adoption. with Yale’s EAD best practices is monitored via external validation against a RelaxNG schema.26 A variety of tools exist to help mask the complexity of EAD and smooth the encoding path. Tools that assist in migration can also aid in the Forms tagging process. See the section on migrating to An EAD finding aid can be presented as a fill- a database or content management system (page in-the-blank HTML form so that the archivist 49). Other tools that can help mask complexity never sees any EAD at all. Although it does are listed in Appendix II. Please note that some of require behind-the-scenes coding to add the these tools are local solutions or strategies—that EAD tags, several institutions have employed this is to say, they have been customized for use within approach and there is likely a Web-based form a consortium, institution or repository and may that your institution can use immediately, or with not work in your setting. We’ve included these minimal adaptations. We’ve included examples because they may provide inspiration for your of several Web-based forms along with two Specialized Migration own use. The section on “ examples of stand-alone forms in Appendix II. and Conversion Tools” may be especially useful. 44 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Style sheets I don’t know where to start!27 Style sheets can display or hide various attributes or other text to enhance readability and aid in completing a template. If using Solution: Take it one step at a time oXygen, XMetaL, or some other commercial and create a plan. The process can XML authoring tool, a Cascading Style Sheet be broken into logical steps. (CSS) file can be automatically created the first time you open an EAD document; this style sheet can then be modified, enhanced, If you are overwhelmed and don’t know how to etc., to assist editors visually. Some examples get started, it may be helpful to think about EAD of stylesheets are included in Appendix II. implementation in terms of a number of small steps. Start with what you have and where you …and more! need to go. Steps usually include documenting We’ve included information about other means inhouse standards (e.g., should extent be of simplifying EAD implementation including expressed as linear feet or cubic feet? Are there pointers to commercial XML tools, content existing best practice guidelines you wish you management systems for archives, and a adhere to? What controlled vocabulary will you variety of papers, production guides and case use for the controlaccess terms), selecting studies. You will find lists in Appendix II. an encoding method (e.g. template with XML editor, database, full content management tool), data entry, selecting or creating a style sheet (to format the XML), and putting the files online. Additional decision areas may include whether and how to provide search capability, whether to provide alternate formats (e.g., a printer-friendly version), whether to link finding aids to digitized content, and so on. For an overview of possible steps, we have included one such plan in Appendix III, and this example may get you started. You may also refer to the EAD Application Guidelines, specifically Chapter 2, for more details and additional food for thought.28

45 Making Archival and Special Collections More Accessible

Section II: Addressing Technical Problems

I want my data to be stored to handle particular situations is available on the Syracuse University Library Web site.31 in a format that will give me flexibility going forward29 In-house encoding In-house conversion offers numerous benefits: fostering staff skills, flexibility in schedule Solution: EAD is a non proprietary and workflow, and direct control over process component of a flexible framework. and inputs. Basic XML skills are not difficult to acquire, and having internal staff with XML knowledge may benefit other library processes In any conversion process, you should be and projects as well. Various tools can speed/ concerned with maintaining flexibility, and not ease creation of EAD. The two following options— losing information. Converting paper files to authoring tools and scripts—assume starting electronic format increases the usability of your with electronic files (e.g., word processing files, or data; EAD encoding offers both flexibility and text files obtained through converting hard copy additional options since EAD can be “crosswalked” using optical character recognition, or OCR). into other standards like MARC XML and Dublin As part of in-house conversion you can use Core. Going further, a well-chosen database or commercial XML authoring tools (we’ve included CMS (Content Management System) can provide a list of the most well-established in Appendix additional output options. It’s all about choices II). You may also be able to manipulate the along this continuum. We’ve presented three text using scripts. Once you have gained ways to think about moving forward—each confidence in understanding and defining the one will advance you to greater flexibility. EAD output you expect, then any programmer with experience of scripting languages like Perl Option 1: Migrating hard copy or or VisualBasic and XML could write scripts to word-processing files to EAD produce the desired output from your existing input. EAD produced this way can be validated Outsource and/or edited using commercial tools, or using 32 Outsourcing data conversion is the fastest and the free online W3C validator or validation easiest solution for hard copy or word-processing could be part of the scripted process(es). file conversion. Depending on your staff costs, this may or may not be the cheapest option. Option 2: Exporting EAD from a Generally conversion houses quote a per-page content management system (for hard copy) or a per-byte (for electronic files) rate. Outsourcing requires considerable up-front Some content management systems (CMS), such work in determining tagging specifications, since as the Archivists’ Toolkit or Archon, are capable of EAD is highly flexible. Adhering to DACS, ISAD(G), producing EAD. If your CMS does not output EAD and/or RLG’s EAD Best Practice Guidelines30 or directly, two key questions are 1) whether required other widely-accepted standards or guidelines is EAD elements are separate fields in the database, strongly advised. Standards and “best practices” and 2) whether your database exports XML. save time and effort, and your end product will be more likely to work with widely-used tools and in aggregations of other EAD files. The EAD Listserv is an excellent resource in terms of getting up to the minute information regarding current vendors. Outsourcing is most successful when both parties are very clear on the encoding specifications. A sample encoding specification, including where to get various pieces of data and how 46 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Table 1. Tips for producing EAD from managed content under various export scenarios.

Data Separate Exports Notes Fields? as XML? Export the data as XML and then use XSLT to convert to EAD. (XSLT is a language for transforming XML documents into HTML documents Yes Yes or to other XML documents, in this case transforming data to EAD). This requires some knowledge of XSL, or the funds to contract out XSL development. Ideally it would be a one-time cost.

Export the data in some other structured form (comma-separated values, for example) and identify a scripting approach to process the data and Yes No convert to EAD. This requires some programming or scripting knowledge. Perl is an example of a scripting language that is useful in this context.

Determine whether scripts can be written to parse output from your database and generate EAD; whether, and if so how, the database No No needs to be modified to be able to export EAD; or whether perhaps the appropriate solution is migration to an EAD-capable database.

Option 3: Migrating to a database or content Making a choice management system capable of producing The open source/commercial distinction is EAD for permanent storage and maintenance one of many factors that should play a role in Using a database to create and store data your archival management system decision- elements of finding aids simplifies data entry, making process. The most important part of reduces the possibility of tagging errors, ensures selecting a system is to choose one that has consistency in output, and offers the possibility the features you need. Resources such as of exporting to formats other than EAD. However, Archival Management Software: A Report for the some full-fledged archival management systems Council on Library and Information Resources may be “overkill” for a legacy conversion project will help you with the selection process. in terms of features, price, and learning curve. If your data is in spreadsheet or word processor Availability of open source software format, or in a database that will not map directly At least two tools that produce EAD are distributed to EAD, migrating to an EAD-capable database as open source software (OSS).34 OSS is produced may be a useful solution. The key question here in a way that allows others to adapt, modify and is whether the data is easily mapped to the redistribute the underlying code and is frequently target database, and whether the time involved associated with a “community” of developers. in migration will in the long run result in the best solution for your needs. A list of content Your institution is most likely using open management systems is included in Appendix II. source software already in some context and that may make it easier to bring in an open source EAD tool. You may be using the Doors are closed to open source33 Apache Web server, database platforms such as MySQL, and desktop applications such as the Mozilla Firefox Web browser. Several Solution: Outline the upsides open source integrated library systems are of open source software available, including Koha and Evergreen35 Open source digital repository systems include Fedora, DSpace, EPrints, and Greenstone.36

47 Making Archival and Special Collections More Accessible

Lack of conflict with commercial Deliver EAD directly to the browser software/commercial enterprises This is by far the simplest and easiest approach. Some institutions have a policy against You can deliver XML directly to most recent Web implementing open source software, preferring browsers (e.g. IE 5+, Firefox .9+). To transform the instead to license or purchase software that EAD XML file to HTML within the Web browser (on includes support or is backed by a reputable the client side), include a processing instruction company. Open source software does not in the XML document pointing either to an XSLT preclude commercial support. Support stylesheet38 (the preferred method) or CSS file.39 contracts are available for many open source However, some institutions may not want to software packages, including the open provide access to their raw XML files, particularly source ILS system previously mentioned. if they include sensitive information in their Commercial support for OSS EAD tools is not finding aids that they don’t want to display to currently available, but this is evolving. the public. Moreover, browser support for XML is still uneven40 (for instance, at the time of the writing of this report, ’s Chrome browser 37 Problems with publishing is reported to not provide full XML support).

Solution: Let the browser do the work, Convert your EAD to HTML or or use existing tools that incorporate PDFs for Web display publication functionality. Instead of displaying the raw XML using a Web browser, convert EAD finding aids to a static files in a human-readable format. By A major obstacle preventing wide-scale adoption applying XSLT stylesheets to XML finding aids, of EAD is delivering EAD-encoded finding aids archives can generate multiple forms of output, to users online. Creating EAD finding aids may including HTML and PDF. Such conversion require a different set of skills than publishing can be accomplished in batch. HTML or PDF them, including authoring XSLT stylesheets, files can then be uploaded to a standard Web installing software, configuring a server, and server to support research and discovery. so forth. There are few inexpensive, “out-of- Developing XSLT stylesheets requires the-box” solutions for publishing EAD online. some technical knowledge, but several However, archives have several options. From consortia and archives have made available simplest to hardest, these include: contributing XSLT stylesheets that archives can easily records to a shared finding aids repository; adapt for their own institutions. Some delivering EAD directly to the browser; converting examples are listed in Appendix II. records to HTML or PDF for Web display; using inexpensive tools to enable searching of HTML Delivering HTML or PDF rather than EAD may be and XML files; using an archival management attractive to archives that lack technical staff system; and using an XML publishing platform. to support XML publishing, but these methods have several drawbacks. They do not take full advantage of having archival information marked Contribute to a shared finding aids repository up in EAD; searches cannot be restricted to Rather than developing their own technical particular EAD elements. Moreover, every time infrastructure for delivering finding aids, some the finding aid is updated, the HTML must be archives choose to deposit them in regional regenerated and uploaded to the server. Some finding aid repositories. The finding aids are archives use a hybrid approach; indexes are hosted centrally and provide a single point of created from EAD files to enable fine-grained access to finding aids from multiple institutions. searching, but the HTML file is delivered to the We’ve included a partial list of finding aid user when they want to view the finding aid. repositories/regional consortia in Appendix I. Syracuse University Libraries take this approach.41 Some archives may want to contribute finding aids to a repository and make them available via their own Web sites.

48 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Use inexpensive tools to enable For the sake of interoperability, the selection searching of HTML and XML files criteria for a commercial archival management system must include the ability to import Even if an archive lacks a substantial budget or and export EAD files, ideally both one at a large technical staff, it can choose from several time and as a batch process. Commercial inexpensive, easy-to-implement tools that packages provided by Adlib, CALM, CuadraStar, support indexing and searching EAD files. One and Eloquent Systems all provide batch and example is Swish-e, “a fast, flexible, and free individual import and export of EAD finding open source system for indexing collections of 42 aids. If your institution requires a hosted Web pages or other files.” Google Site Search solution, many vendors offer such an option. also provides an inexpensive, customizable way of searching your Web pages.43 Use an XML publishing platform Use an archival management system XML publishing platforms enable documents that supports publication to be searched, browsed and displayed. Implementing them requires fairly Many archival management systems enable sophisticated systems administration and publication via export of finding aids in EAD, programming knowledge. Some XML publishing HTML, or PDF. By using archival management platforms are listed in Appendix II. systems, archivists can streamline workflows, avoid duplicating data in multiple places, find and share information more easily, manage collections, and generate reports and statistics.44 A list of archival management systems that support Web publishing of finding aids are listed in Appendix II. Archival management systems have some drawbacks: they may enforce a rigid workflow, it can be difficult to import data, and some are costly to implement. On the other hand, archival management systems can enable archives to create, manage, and share archival information more efficiently.

49 Making Archival and Special Collections More Accessible

EAD can be complicated minimum-level description elements can simplify EAD immensely. (but there’s hope…) EAD was designed to be flexible in order to accommodate a broad range of archival Getting through it practice. In offering flexibility, the standard Despite a more than a decade of practice, has succeeded almost too well. archivists still encounter significant barriers in Freedom of choice in implementation means, EAD implementation. We hope this paper gives for example, that three people could encode the you options to get over hurdles, under obstacles, extent of a collection in three different ways. around complexity, and through difficulty. This flexibility in implementation can cause We recognize that EAD can be challenging. difficulties for aggregators who harvest EAD Examples of EAD’s complexity can be found data from multiple institutions for indexing and easily by looking through the EAD Tag Library. searching. It also hinders tool development Many elements, including accessrestrict, since tool builders must either allow for multiple controlaccess, bioghist, and note, may encoding options or choose one “right way,” be repeated within an element with the same when there are multiple correct ways to encode name to an arbitrary depth; for example, EAD the same thing. EAD’s inherent complexity allows one to encode nested controlaccess makes it difficult for institutions to make elements with no restrictions on how deep that decisions regarding implementation. Those who nesting goes. In addition, EAD has seventeen are choosing tools must evaluate the choices linking elements; of those seventeen, twelve of made by tool builders to ensure that outputs those elements allow the href attribute, which meet their own best practice guidelines. allows linking to resources external to a given So what to do? Make a decision. Document the EAD file. Elements that allow “mixed content” decision. Apply it consistently. Until the flexibility (those that can contain both text and other inherent in EAD is in reigned in, institutions can elements in arbitrary order) can present problems maximize the consistency of their data by: when importing EAD to a database or porting to another data scheme. Some elements that can 1) Selecting a template in use at one or more be full of mixed content and contain information institutions, or creating a template that that would be lost in migration to a database (or adheres to a “best practice” document in would require additional tagging after export) use at one or more than one institution. are p, listitem, bibref, and head. Once you’ve established a template, deviate from it as little as possible. EAD will be under active review in the near future. We recommend that the Technical 2) Clearly document how dates, extent, Subcommittee for EAD (the soon-to-be charged etc., should be encoded. Follow your successor to the EAD Working Group) and own documentation rigorously. the archival community as a whole consider 3) Refrain from excessively complex ways in which EAD can be simplified. coding (for example, nesting As reflected in the large number of EAD duplicate scopecontent tools listed in this paper and its appendices, elements within each other). there are many choices for would-be EAD 4) Refrain from adding unnecessarily implementers. While diversity and choice is a good thing, the range and number of available elements such as list elements within a control access simply to achieve a desired tool choices provide additional complexity. By appearance in the output. EAD should highlighting tools that are already available, we be only be used to encode the structure encourage institutions to utilize work that has and content of a document; appearance been done elsewhere and not to invest what should be controlled by the stylesheet. might be unnecessary development effort. Remember, too, that the entire EAD tagset need not be used. As mentioned above, limiting yourself to collection-level descriptions and the DACS

50 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Acknowledgments We are grateful to the reviewers who provided valuable and useful feedback on early drafts of this document: • Doug Dunlop Smithsonian Institution • Glenn Gardner Library of Congress • Suzanne Pilsk Smithsonian Institution • Michael Rush Yale University • S. Diane Shaw Smithsonian Institution • Bradley D. Westbrook University of California, San Diego • Jackie Dooley OCLC Research • Ricky Erway OCLC Research • Jennifer Schaffner OCLC Research

51 Making Archival and Special Collections More Accessible

Appendix I. Consortia and EAD Aggregators

Almost all can provide a means of publishing finding aids, or may serve as an additional distribution channel for collection descriptions. Many also have tools to aid in EAD creation, provide instruction opportunities, and have developed best practice guidelines.

United States UK and Continental Europe • Archival Resources in Wisconsin: • A2A (Access to Archives, United Kingdom): http://digital.library.wisc. http://www.nationalarchives.gov.uk/a2a/ edu/1711.dl/wiarchives • Archives Hub (United Kingdom): • Archives Florida: http://www.archiveshub.ac.uk/ http://palmm2.fcla.edu/afl/ • Archives Portal Europe: • Arizona Archives Online: http://www.apenet.eu/ [forthcoming] http://azarchivesonline.org • MALVINE (Europe): • Black Metropolis Resources Consortium: http://www.malvine.org/ http://www.blackmetropolisresearch. • National Archival Database of Sweden: org/ [forthcoming] http://nad.ra.se/static/back_eng.html • Historic : http://digital.library.pitt.edu/pittsburgh/ Subject based • Kentuckiana : http://kdl.kyvl.org/ • Navigational Aids for the History of Science in Europe (NAHSTE): • Digital Library: http://www.nahste.ac.uk/ http://msdiglib.net/ • Guide to Australian Literary Manuscripts: • Mountain West Digital Library: http://findaid.library.uwa.edu.au/ http://mwdl.org/index.php/ search/results?format=ead • Irish Literature Collections Portal: http://irishliterature.library.emory.edu/ • Northwest Digital Archives: http://nwda.wsulibs.wsu.edu • Physics History Finding Aids Web site (PHFAWS): • OhioLINK: http://ead.ohiolink.edu http://www.aip.org/history/ • Online Archive of California (OAC): nbl/findingaids.html http://www.oac.cdlib.org/ • Pennsylvania Digital Library: Other http://padl.pitt.edu/ • ArchiveGrid: • Rhode Island Archival and Manuscript http://www.archivegrid.org Collections Online (RIAMCO): OCLC’s Archive Grid combines finding aids www.riamco.org [forthcoming] with MARC records to create one-stop- • Rocky Mountain Online Archive: shopping for users. Heavy representation http://rmoa.unm.edu from US institutions, also representation from outside the US. Contribution is • Texas Archival Resources Online (TARO): free and open to any institution. http://www.lib.utexas.edu/taro • Virginia Heritage Project: http://www2.lib.virginia.edu/small/vhp/

52 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Appendix II. Tools

Templates • Notre Dame: http://classic. archives.nd.edu/ead/ead.htm Examples of templates include: includes both collection level • Northwestern University: and inventory forms http://staffweb.library.northwestern. • Berkeley: http://sunsite3.berkeley. edu/dl/ead/template.xml edu/ead/tools/template/ • Syracuse University: cgi Web application; appears to be collection- http://library.syr.edu/digital/ level only. Last update to page is 2005. guides/ead/aaa_template.xml • Western Kentucky: http://pax.uky. • Indiana University Bloomington: edu/template-v1-cgi/template. http://www.dlib.indiana.edu/services/ pl/KNVUA_generic.tmp metadata/activities/eadDocumentation.shtml • Austin College Xforms tool: http://www. archivists.org/saagroups/ead/tools.html Web-based forms Developed by Justin Banks at Austin Following are a few examples of Web-based College. Requires server that supports forms. Some produce only collection-level Xforms. Additional information from SAA EAD (that is, they do not include the inventory 2007 available here: http://matienzo. in the section), while others produce org/saa2007descriptionexpo. complete inventories. They differ in ease of use, complexity, and quality of product. The first two are probably the most complete. Standalone forms • California Digital Library: http:// • The University of has a java-based EAD www.cdlib.org/services/dsc/tools/ authoring tool called xEAD, currently publicly ead_webtemplates.html available at https://lsta.lib.byu.edu/lstawiki/ index.php/XEAD_Project. The application The California Digital Library has created opens an EAD file in its buffer, allowing users numerous online templates that create to manipulate the data and then resave. EAD from typed or copy-pasted data. One or more of these may be adaptable • The German Bundesarchiv, with assistance for your institution’s needs. from the Mellon Foundation, has developed a tool called Midosa Editor for XML or MEX. • ArchivesHubUK: http://www. It is available in English and German and archiveshub.ac.uk/eadform2002.html for both OSX and Windows. MEX provides Web-based form, renders complete EAD an authoring environment for creating a document. Enables editing of uploaded variety of levels of EAD records and includes files, creation of new ones, saving of draft built-in publication to HTML capability. file between editing sessions, preview Background information is available at feature. Links to digital surrogates are http://www.bundesarchiv.de/daofind/ easy to add, common markup tags (e.g. en. Downloads and a quick-start guide are paragraphs, lists, titles, links) can be available from the MEX SourceForge wiki at added via a right-click menu, and a special http://mextoolset.wiki.sourceforge.net/. characters keyboard is provided. Components can easily be added to create sub- descriptions, and a tree structure will show Style sheets for authoring exactly what the hierarchy looks like. finding aids • Northwestern University: http:// • Yale University has developed the Yale staffweb.library.northwestern.edu/dl/ Finding Aids Creation Tool (FACT), which is ead/eadchef/template.cgi/ead/nul a customized version of XMetaL Author. The

53 Making Archival and Special Collections More Accessible

tool includes a set of style sheets, macros, Content Management and other customizations. More information is available at http://yalefact.pbworks.com. Systems for Archives • Northwestern Digital Archives: http:// • Archon: http://www.archon.org nwda.wsulibs.wsu.edu/tools/ead.css Open source full archival management • Syracuse University: http://library. system, developed by University of Illinois syr.edu/digital/guides/ead/ead.css with funding from Mellon Foundation. Capable of ingesting MARC or CSV format. Can export EAD and MARC. XSLT stylesheets for • Archivists’ Toolkit: http://www. displaying finding aids archiviststoolkit.org • EAD 2002 Cookbook: http://www. Open source full archival management archivists.org/saagroups/ead/ system, developed by UCSD, NYU, and ead2002cookbookhelp.html Five Colleges Inc. and funded by Mellon Foundation. Capable of ingesting tab- • EAD Help pages’ user contributed style delimited fields and MARC as well sheets: http://www.archivists.org/ as preexisting EAD. Can export EAD, saagroups/ead/stylesheets.html Marc, Dublin Core, MODS, METS. • UC Berkeley EAD Tools: http:// • MSAccess, FileMakerPro, etc. sunsite3.berkeley.edu/ead/tools/ Commercial but reasonably priced general • NC Echo EAD Tools: http://www. database development tool. A database ncecho.org/dig/ncead.shtml#tools developed in-house with the appropriate • University of Minnesota: https://wiki. fields, in conjunction with XSL, java, or other lib.umn.edu/Staff/FindingAidsInEAD scripting languages, can generate EAD.

Commercial XML tools Archival management systems for EAD encoding that support publishing • oXygen: http://www.oxygenxml.com/ • Adlib (commercial): http://www.adlibsoft.com/ Commercial XML authoring and • Archon (open source): http://archon.org/ editing software. Data entry, copy- • Calm (commercial): http://www. paste; can handle DTD or schema crxnet.com/page.asp?id=57 • XMetaL: http://na.justsystems. • Cuadra STAR/Archives (commercial): http:// com/content-xmetal www.cuadra.com/products/archives.html Commercial XML authoring and • Eloquent (commercial): http://www.eloquent- editing software. Data entry, copy- systems.com/products/archives.shtml paste; can handle DTD or schema. • ICA-AtoM (open source): http://ica-atom. • XMLSpy: http://www.altova.com org/ [currently available in Beta release] Commercial XML editing, authoring, • Minisis M2A (commercial): http://www. development environment. minisisinc.com/index.php?page=m2a • NoteTab Pro: http://www.notetab.com/ General purpose text editor which can be Various other commercial archival easily customized to handle EAD. Clip libraries and other add-ons are available here http:// management systems can import www.archivists.org/saagroups/ead/tools.html and export EAD, including: • Minisis: http://www.MINISISinc.com

54 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

• Eloquent Archives: http://www.eloquent- • DLXS: http://www.dlxs.org systems.com/products/archives.shtml XML-aware search engine (XPAT or XPAT Lite) • AdLib: http://www.adlibsoft.com with DLXS middleware which includes a “class” for finding aids (currently in prototype). • CALM: http://www.ds.co.uk (Open source component, with For detailed information on these and commercial options available; other commercial tools, refer to Lisa Spiro’s used by University of Michigan and 2009 report for CLIR, Archival Management University of Minnesota Libraries46) Software http://clir.org/pubs/reports/( spiro2009.html), or to the associated wiki • Cheshire3: http://www.cheshire3.org/ at http://archivalsoftware.pbwiki.com/. “Fast XML search engine.” Standards compliant, with support for Open XML publishing platforms45 Access Initiative (OAI) protocols and Z39.50. Modular and configurable.” • XTF: http://xtf.wiki.sourceforge.net/ (Open source, used by University of “A flexible indexing and query tool that Liverpool Special Collections and supports searching across collections of Archives and ArchivesHub) heterogeneous data and presents results in a highly configurable manner.” XTF supports powerful searching, faceted browsing, Specialized migration and viewing search terms in context. or conversion tools (Open source; used by California Digital • MARC to EAD—MarcEdit: http://oregonstate. Library and numerous others) edu/~reeset/marcedit/html/index.php • Mark Logic: http://www.marklogic.com/ If you have MARC records for your manuscript The University of Chicago is developing collections, you can quickly and easily an XML publishing infrastructure built on generate skeleton (i.e., collection-level) EAD MarkLogic, a native XML database. The front records from it using MarcEdit. MarcEdit end can be built on any platform and provides uses xsl style sheets which can easily be flexible display options. The University of modified/customized. Developed by Terry Chicago’s code will be available to anyone. Reese at State University; free. Archives that want to use the software will • Excel to EAD using Mailmerge need MarkLogic, but there is a free (limited) version that will suffice for small institutions. Excel is an immensely useful tool for generating the code for lengthy inventories (Commercial; used by University of Chicago) of minimal depth. Text can be entered into • PLEADE: http://www.pleade.org/en/ a spreadsheet, then columns can be added before and after the text and populated with “open source search engine and browser the correct EAD elements. For collections with for archival finding aids encoded in XML/ large inventories, either copy-pasting (from EAD. Based on the SDX platform, it is Word, RTF, or txt) or entering afresh in Excel a very flexible Web application.” may be a workable solution. Excel’s MailMerge (Open source; used by Denver ) feature can also be employed to automatically • Cocoon: http://cocoon.apache.org/ generate coded data from an Excel spreadsheet; see video here http://archives. Cocoon is an open source XML publishing state.ut.us/containerlist/containerlist.html. framework that applies XSLT stylesheets This could be used in conjunction with to the EAD finding aid to display MarcEdit which generates the collection- HTML. Used in tandem with indexing level part of EAD to produce a full EAD technologies such as Lucene or eXist. inventory. Indiana University has posted (Open source; used by Ohio State and the detailed instructions and an Excel template Five College Finding Aids Access Project) for encoding lengthy inventories at http:// www.dlib.indiana.edu/services/metadata/

55 Making Archival and Special Collections More Accessible

activities/eadDocumentation.shtml (see the – Draft of paper looking at steps in an EAD section “Using Excel to assist with encoding”). conversion project; very rough but good info. • Text to EAD—EAD Conversion: • University of Indiana EAD guide: http:// http://agileimage.com/html/ead/ www.dlib.indiana.edu/services/metadata/ activities/EADManual.pdf – includes Reads a text version of inventory and detailed instructions for using oXygen. generates an EAD-encoded version. Last update to Web site was in 2004. • Utah State Archives EAD Project: http:// www.archives.state.ut.us/research/ • Text to EAD—MSWord inventories/ead.html – includes detailed Starting with an electronic file of an discussion of how they converted their inventory, a surprising amount of tagging legacy finding aids using a combination of can be done simply using MSWord’s tools (HTML, Excel, WordPerfect, etc). search and replace feature with tabs and • NYU Archives EAD Production Guide: regular expressions, including locating http://www.nyu.edu/library/bobst/ and tagging , , long unitdate extent research/arch/eadProduction.htm – inventories, etc. Good overviews of regular detailed procedures used by NYU, includes expressions in word are available here: detailed instructions for NoteTab Pro. —— http://office.microsoft.com/en-us/ • The EAD Help Pages: http://archivists.org/ help/HA010873051033.aspx saagroups/ead/sitesann.htm – more than —— http://office.microsoft.com/en-us/ 80 institutions currently implementing help/HA010873041033.aspx EAD, including brief descriptions of each —— http://word.mvps.org/FAQs/ institution’s approach and a point-of-contact. General/UsingWildcards.htm Implementations run the gamut from extremely simple (EAD put online with a style • Date normalization—tri-XMLdate- sheet) to extremely sophisticated (databases normalizer.pl: http://www.archivists. that provide server-side transformations org/saagroups/ead/tools.html and advanced search capabilities). Developed by Jason Casden at The ; free. For large files, inserting the normal attribute for unitdate elements can be extremely time-consuming. This Perl script automatically recognizes numerous date formats and inserts the normal attribute. Offers options to overwrite existing values or leave them.

Papers, production guides, case studies, etc. • Northwest Digital Archives Standards Working Group Review of Web Templates for EAD Creation (October 2008): http:// www.orbiscascade.org/index/cms- filesystem-action?file=nwda/files/ bowmanreview_200810.pdf – excellent and thorough review of available Web- based templates that generate EAD. • OhioLINK EAD Starter Package (April 2008): http://platinum.ohiolink.edu/dms/ead/ contentguide/EAD_starter_packet_v4.pdf

56 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Appendix III. EAD Migration, Creation and Publication Paths

Figure 1. EAD Migration and/or Creation

57 Making Archival and Special Collections More Accessible

Figure 2. EAD Publishing Paths

58 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

Simplest path (conversion from MARC records) • Export collection-level EAD records from MARC using MarcEdit. • Validate and correct errors using W3C’s online validation tool (http://validator.w3.org ). • Select or create an XSL style sheet. • Put files on server.

Simplest path (starting without MARC) • Use one of the many existing Web- based templates to generate collection-level EAD files. • Validate and correct errors using W3C’s online validation tool (http://validator.w3.org ). • Find or create an XSL style sheet. • Put files on server. The above solutions require minimal time, expertise, and money, yet yield online collection- level descriptions that will be “crawled” by Google and other search engines making them discoverable via the open Web, and valid EAD files which can be contributed to consortia (a list of consortia can be found in Appendix I) or aggregators like OCLC’s ArchiveGrid. Files can easily be “upgraded” at some later date without reworking.

More sophisticated path • Choose one of the full collection-management packages such as Archivists’ Toolkit or Archon. • Perform data entry and/or import to level desired. • Install and configure XTF installation for Web availability, browsing, searching, etc. • Export EAD to be indexed/ searched by XTF system. This yields a fully-functioning database and sophisticated search capabilities but requires substantial technical knowledge to install and configure the XTF installation.

59 Making Archival and Special Collections More Accessible

Notes

All Web links were verified 25 February 2010. 7 7. Yaco, “It’s Complicated,” 468: “another key barrier is ‘Lack of institutional 1 EAD Roundtable, EAD (Encoded Archival support’.”; Wisser, EAD Tools Survey, 19: Description) Help pages, sponsored “Some folks in the organization… are by the Society of American Archivists, unconvinced of EAD’s staying power and http://www.archivists.org/saagroups/ thus stingy when it comes to allocating ead/. In particular, see the section on resources and time for relevant training. “What is EAD,” at http://www.archivists. Is there an EAD eye-opener kit?” org/saagroups/ead/aboutEAD.html. 8 International Council on Archives. 2 Archivists Toolkit User Group, “2008 ISAD(G): General International Standard AT User Group Survey Results,” http:// Archival Description, Second Edition, archiviststoolkit.org/sites/default/ http://www.ica.org/en/node/30000; files/AT%20User%20Group%20 Society of American Archivists, SurveyResultsFD.pdf. Additional Describing Archives: A Content Standard breakdown of survey results supplied by (DACS), http://www.archivists.org/ Brad Westbrook to the working group governance/standards/dacs.asp. via e-mail on 18 December 2008. 9 Society of American Archivists. “Appendix 3 Tatem, Jill, “EAD: Obstacles to A: EAD Crosswalks,” Encoded Archival Implementation, Opportunities for Description Tag Library, Version 2002, Understanding,” Archival Issues 23,2, The Library of Congress, http://www. (1998): 155-169; Roth, James M., “Serving loc.gov/ead/tglib/appendix_a.html. Up EAD: An Exploratory Study on the Deployment and Utilization of Encoded 10 Spiro, Archival Management Software. Archival Description Finding Aids,” 11 U.S. Library of Congress Network The American Archivist, 64,2 (2001): Development and MARC Standards 214-237; Dow, Elizabeth H, “EAD and Office, EAD (Encoded Archival Description) the Small Repository,” The American Electronic List, http://www.loc.gov/ Archivist, 60,4 (Fall 1997): 446-455. ead/eadlist.html; Society of American 4 Wisser, Katherine M., EAD Tools Survey, Archivists, Web site, www.archivists. Society of American Archivists, EAD org; Rare Book School, Web site, Roundtable, (August 2005), http:// http://www.rarebookschool.org/. www.archivists.org/saagroups/ead/ 12 Szary, Richard V., “Encoded Finding EADToolsSurvey.pdf; Yaco, Sonia, Aids as a Transformative Technology “It’s Complicated: Barriers to EAD in Archival Reference Service,” in Implementation,” The American Archivist, Encoded Archival Description on the 71,2 (Fall/Winter 2008): 456-475. Internet, ed. Daniel V. Pitti and Wendy 5 A recent report from CLIR documents the M. Duff, 187-197. Binghamton, NY: current state of archival management Haworth Information Press (2001). software, much of which provide support 13 ArchivesZ, http://archivesz.com/; Syracuse for EAD: Lisa Spiro, Archival Management University Library, “Elastic Lists: EAD Software, A Report for the Council on demo”, http://library.syr.edu/digital/ Library and Information Resources, CLIR guides/ElasticLists-EAD/binv3/index.html; Reports (January 2009), http://www. NNDB Mapper, http://mapper.nndb.com/. clir.org/pubs/reports/spiro2009.html. 14 National Endowment for the Humanities, 6 Children’s Television Workshop, Sesame Division of Preservation and Access, Street, Episode 0299 (December 9, 1971). “Frequently Asked Questions: Humanities Collections and Resources,” http:// www.neh.gov/grants/guidelines/ collections&resfaqs.html. 60 Over, Under, Around and Through: Getting Around Barriers to EAD Implementation

15 National Information Standards 24 “MarcEdit—Your Complete Free MARC Organization, A Framework of Guidance Editing Utility,” (Web site hosted for Building Good Digital Collections, 3rd at Oregon State University) http:// Edition, with support from the Institute people.oregonstate.edu/~reeset/ for Museum and Library Services (IMLS), marcedit/html/index.php. http://framework.niso.org/node/38. 25 Yaco, “It’s Complicated,” 460: 16 National Historical Publications and “the complexity of EAD itself [is] a Records Commission (NHPRC), “Basic deterrent to implementation.” Projects Grant Announcement,” http:// 26 Yale University Library FACC: Finding www.archives.gov/nhprc/announcement/ Aids Coordinating Committee, basic.html, and “Detailed Processing “Yale University EAD Encoding Best Projects Grant Announcement,” Practice Guidelines,” http://www. http://www.archives.gov/nhprc/ library.yale.edu/facc/bpgs.html. announcement/detailed.html. 27 Yaco, “It’s Complicated,” 17 EAD Roundtable, “90 Implementors 466: “Lack of a plan”. Currently Listed,” EAD Help Pages http://www.archivists.org/saagroups/ 28 Society of American Archivists, Encoded ead/implementors.html. Archival Description Working Group, Encoded Archival Description Application 18 Yaco, “It’s Complicated,” 459: Guidelines for Version 1.0, (Encoded “Concern about the quality or Archival Description (EAD), Document completeness of finding aids often Type Definition (DTD), Version 1.0, causes archivists to rewrite legacy Technical Document No. 3), (1999), http:// finding aids before EAD encoding.” www.loc.gov/ead/ag/aghome.html. 19 Voltaire, La Bégueule: Conte moral, (1772). 29 Spiro, Archival Management Software, 6: 20 Mark A. Greene and Dennis Meissner, “Archival material is so specific that you More Product, Less Process: Pragmatically don’t want to get locked into anything… Revamping Traditional Processing Ideally, I would want something that Approaches to Deal with Late 20th- would also preserve that information in a Century Collections, http://ahc.uwyo. format that is able to migrate if needed.” edu/documents/faculty/greene/ 30 RLG EAD Advisory Group, RLG EAD papers/Greene-Meissner.pdf, page Best Practice Guidelines for Encoded 2. (Also published as “More Product, Archival Description, (August 2002), Less Process: Revamping Traditional http://www.oclc.org/programs/ Archival Processing,” American ourwork/past/ead/bpg.pdf. Archivist 68,2 (2005): 208-63.) 31 Michele Combs, EAD Tagging Specs 21 Society of American Archivists, (Revised 10-3-08), Syracuse University “Code of Ethics for Archivists,” http:// Library, http://library.syr.edu/digital/ www.archivists.org/governance/ guides/ead/tagging_specs.doc. handbook/app_ethics.asp. 32 World wWde Web Consortium 22 Society of American Archivists, “ALA- (W3C), Markup Validation Service, SAA Joint Statement on Access: http://validator.w3.org. Guidelines for Access to Original Research Materials,” http://www. 33 Spiro, Archival Management Software, archivists.org/statements/alasaa.asp. 12: “Some institutions, however, lack the technical staff to implement 23 Yaco, “It’s Complicated,” 466: “lack open source software. Others may of staff...[along with] lack of time.” oppose it because of they fear security risks or high maintenance costs.”

61 Making Archival and Special Collections More Accessible

34 Archon and the Archivists’ Toolkit are both open source applications. At this writing, the two groups are looking into the possibility of combining the best features and functionality of each application into a single application.

35 Koha, http://www.koha.org/; Evergreen, http://open-ils.org/.

36 Fedora, http://fedora-commons.org/; DSpace, http://dspace.org/; EPrints, http://eprints.org/; Greenstone, http://www.greenstone.org/.

37 Yaco, “It’s Complicated,” 461: “Even archives with the resources to create custom, sophisticated encoding computer programs struggle with publishing issues.”

38 Refsnes Data, w3schools.com, “Displaying XML with XSLT,” XML Basic, http://www. w3schools.com/Xml/xml_xsl.asp.

39 Refsnes Data, w3schools.com, “Displaying XML with CSS,” XML Basic, http://www. w3schools.com/Xml/xml_display.asp.

40 Bob DuCharme, “Using XSLT to Deliver XML on Browsers,” bobdc.blog, http:// www.snee.com/bobdc.blog/2008/09/ using-xslt-to-deliver-xml-on-b.html.

41 Syracuse University Library, Special Collections Research Center, http://library. syr.edu/information/spcollections/.

42 Swish-e, http://swish-e.org/.

43 Google Inc., Google Site Search, http:// www.google.com/sitesearch/.

44 For more information on archival management systems, see the Archival Software wiki, http://archivalsoftware. pbwiki.com/FrontPage.

45 This section adapted from Spiro, Archival Management Software.

46 University of Minnesota Libraries, “Finding Aids in EAD,” Libraries Staff Wiki, https:// wiki.lib.umn.edu/Staff/FindingAidsInEAD.

62 Making Archival and Special Collections More Accessible 6 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid, and Implications for Discovery Systems

By M. Bron, M. Proffitt and B. Washburn

This paper was originally published in Code4Lib Journal, 22 (2013-10-14) at http://journal.code4lib.org/articles/8956.

63 Making Archival and Special Collections More Accessible

The ArchiveGrid discovery system is made up in part of an aggregation of EAD (Encoded Archival Description) encoded finding aids from hundreds of contributing institutions. In creating the ArchiveGrid discovery interface, the OCLC Research project team has long wrestled with what we can reasonably do with the large (120,000+) corpus of EAD documents. This paper presents an analysis of the EAD documents (the largest analysis of EAD documents to date). The analysis is paired with an evaluation of how well the documents support various aspects of online discovery. The paper also establishes a framework for thresholds of completeness and consistency to evaluate the results. We find that, while the EAD standard and encoding practices have not offered support for all aspects of online discovery, especially in a large and heterogeneous aggregation of EAD documents, current trends suggest that the of the EAD standard and the shift from retrospective conversion to new shared tools for improved encoding hold real promise for the future.

different possible dimensions of a discovery Introduction system: search, browse, sort, limit, and display. ArchiveGrid is an aggregation of nearly two million As a warning to the reader: this paper delves archival material descriptions, including MARC deeply into EAD elements and attributes records from WorldCat and finding aids harvested and assumes at least a passing knowledge from the web. It is supported by OCLC Research of the encoding standard. For those wishing as a corpus for experimentation and testing in to learn more about the definitions and text mining, data analysis, and discovery system structure, we recommend the official applications and interfaces. Archival collections EAD website or the less official but highly held by thousands of libraries, museums, readable and helpful EADiva site2. historical societies, and archives are represented in ArchiveGrid. Although roughly 90% of what is in ArchiveGrid are MARC records, as of April Related Work 2013 OCLC Research had harvested 124,009 EAD encoded finding aids for inclusion in ArchiveGrid1. The work that is the most closely related to our This small segment of ArchiveGrid is important research was done by Katherine M. Wisser and because EAD has been embraced by the archival Jackie Dean1. In 2010 Wisser and Dean solicited community since it’s inception in the 1990s, EAD files repositories from institutions in order and is supported by a range of tools designed to ”identify encoding behavior.”3 In total, 108 specifically for archives, such as ArchivesSpace, repositories submitted up to 15 finding aids Archivists’ Toolkit, Archon, CALM, and others. for the analysis; 1,136 finding aids comprise the entire sample. The formal results of their In creating the ArchiveGrid discovery interface, analysis will be published in the Fall 2013 edition the project team has wrestled with what we can of American Archivist. We are grateful to the reasonably do with this corpus. For example, authors for sharing their early work with us, it would be useful to be able to sort by size of and note with interest that in many cases, their collection, however, this would require some analysis of EAD usage is quite similar. However, level of confidence that the tag is in some notable cases, the findings from the two both widely used and that the content of the tag samples diverge dramatically (see for example would lends itself to sorting. Other examples of elements in above the desired functionality include providing a means in Table 9). As noted by Wisser and Dean some in the interface to limit a search to include only of this variation can be attributed to the many items that are in a certain genre (for example, different ways in which EAD files can be obtained. photographs) or to limit a search by date. Wisser and Dean invited a limited contribution Again, we would need to have confidence that (12-15 finding aids) from a wide variety of the metadata we have will actually support repositories, including significant contributions these features, and not leave out potentially from institutions outside of the US; even though important collections simply because of the Wisser and Dean carefully articulated that results absence of certain tags. Specifically, we will would be anonymized, there is some chance consider how the variability of use of elements that the results were somewhat skewed by the in finding aids affects discovery considering five process of selecting files for inclusion. By contrast,

64 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid our data set was assembled by harvesting EAD Methods documents from institutions directly, see below. Contributing institutions have been motivated to contribute to ArchiveGrid primarily to share Defining Thresholds information about their collections, not their It is difficult to predefine thresholds for the level EAD practices. Additionally, ArchiveGrid is of usage of an element at which it becomes primarily constituted by repositories from more or less useful for discovery. Is an element the United States, with few institutions from that is used 95% of the time still useful but Europe or elsewhere represented in the data one that is used 94% not? In this paper we set. Either or both of these key differences may consider the thresholds resulting from working account for divergence in findings between with our sample of documents. We will use our work and that of Wisser and Dean. the terminology documents and finding aids The 2010 report, “Implications of MARC Tag interchangeably throughout the paper. Usage on Library Metadata Practices” focused As an indicator for usage of an element we use on an analysis of the MARC standard as reflected the percentage of documents that contain the in World-Cat5. Although the emphasis of the element at least once (% uniq). The nested nature report was, similar to Dean and Wisser, meant of finding aids, however, influences the usage to “inform community practice,” a secondary of elements as the absence of a parent element purpose was to draw conclusions about the reduces the possibility of the occurrence of child suitability of MARC data for machine matching elements. As an alternative indicator for usage we and processing, which is similar to our desire to use the percentage of documents that contain an identify functionality (and gaps in functionality) element in the sample of documents that contain that exist in our current EAD corpus. the element’s parent element (% uniq in C). OCLC Research regularly harvests EAD Figure 1 shows how often the percentage of usage documents from contributing institutions to of an element falls into certain intervals. Note update their representation in the ArchiveGrid that we use relative usage (% uniq in C) here. index. The update cycle is roughly every six weeks. Institutions are contacted to obtain The distribution of element usage could be their permission to harvest and use the data in roughly divided into 4 groups: (i) usage between ArchiveGrid, and to identify the target URLs and 0%-50% or low use; (ii) usage between 51%-80% rules for selection. For some contributors, the or medium use; (iii) usage between 81%-95% harvesting rules are simple: a directory listing or or high use; (iv) usage between 96%-100% or an HTML page is made available to our crawler, complete use. Although we will use these levels with every link leading to an EAD XML file on the as a reference point in this document, we do so contributor’s server. For other contributors we with a recognition that correlating usage with may make use of a website designed for human discovery is an artificial construct. In the absence visitors, applying custom include and exclude of a more effective approach, we are using these rules to the URLs we find to select only links to levels as an initial framework for discussion. EAD documents. Though OAI-PMH repositories The absence of an element does not directly and other more specialized harvesting protocols lead to a breakdown in a discovery system. It is may be available at some contributor sites, we more like a gradual decay of the effectiveness have seen little interest among contributors of a discovery system. But not all elements are in their use, and currently we are using only created equally – in current archival discovery standard HTTP GET requests for all the many systems, we see a range of functionality that is hundreds of EAD document providers. Maintaining offered, both in terms of search and advanced the EAD harvesting operation continues to be search options, as well as sorting features, and a significant component of the ArchiveGrid results display. Within smaller aggregations, support costs covered by OCLC Research. we might very well expect tag usage to be considerably more internally consistent than is the case in the ArchiveGrid aggregation. But in the case of ArchiveGrid and similar large aggregations of finding aids, what functionality can be reasonably supported, given the present 65 Making Archival and Special Collections More Accessible state of the data? What functionality can we offer with assurance, if we look only at elements that are in the high or complete categories?

Figure 1: The distribution of percentage of element usage (% uniq in C). Elements are nested and the absence of a parent element influences the occurrence percentage of a child element. For this reason we use the number of element occurrences relative to the occurrences of the parent element (% uniq in C).

66 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

Counting Element Occurrences • search: all discovery systems have a keyword search function; many also include the ability Finding aids follow the Encoded Archival to search by a particular field or element: Description standard, which is a complex XML examples include name, date, subject. structure. As an example of the complexity of EAD in implementation, we found more than 26,000 • browse: many discovery systems paths in our 129,009 document set. To provide a include the ability to browse finding starting point for obtaining element counts we aids: examples include browse by recreated the many (but not all) tables of element, repository, browse by material type. attribute, and value counts as presented in the • results display: once a user has done a search, report by Wisser et al.4 Each table was recreated the results display will return portions of the by performing one or more XPath queries over finding aid to help with further evaluation: the corpus of finding aids. In the discussion of our examples include title, dates, collection size. analysis we do not follow the same structure as in Wisser et al.4 as our focus is on implications of • sort: once a user has done a search, element usage on discovery and presentation. they may have the option to reorder Where appropriate similarities and differences the results. Examples include: order by between element usage in our sample of finding date, order by title, order by size. aids and those used in Wisser et al.4 are reported. • limit by: once a user has done a search, they In the rest of the paper we use the following may have the option to narrow the results to notation in our tables: (i) N is the total number only include results that meet certain criteria. of occurrences of an element; (ii) N uniq is the This may be done through presentation of number of documents in which the element facets: examples include limit by collections with digital material, limit by repository. Nuniq occurs at least once; (iii) is the S percentage of documents in our sample of EAD documents (S= 124009) that contain the Current discovery interfaces Nuniq We reviewed a number of different discovery element at least once; and (iv) is the interfaces for finding aids in order to provide n=... percentage of documents that contain the an overview of the type of search, browse, element in the sample of documents (n=…) that sort, limit, and display options that are contain a certain element. We will provide the generally available. Interfaces included are: size of each particular sample explicitly. For • the Online Archive of California example, when considering the (http://www.oac.cdlib.org/), element that occurs in every document we get • the Northwest Digital Archive (http:// Nuniq Nuniq , which is the same as . n=124009’ S nwda.orbiscascade.org/), Nuniqk • Texas Archival Resources Online (http:// We use to indicate the percentage of n=... www.lib.utexas.edu/taro/index.html), documents that contain the element in the sample of documents that contain a certain • Arizona Archives Online (http://www. element as collected by Wisser. In most cases azarchivesonline.org/xtf/search), the sample size will be all documents in Wisser’s • the Five Colleges Archives and Nuniqk Manuscripts Collection (http://asteria. sample, i.e., · . Finally, we use diff n=1136 fivecolleges.edu/index.html), to indicate the percentage point difference between the percentage Nuniq and Nuniqk, • the Rocky Mountain Online Archive i.e., between Wisser’s and our sample. (http://rmoa.unm.edu/), • the Harvard Library’s Online Archival Search Information System (http://oasis.lib.harvard. Dimensions for Analysis edu/oasis/deliver/home?_collection=oasis). Our analysis considered the following dimensions: The interfaces we surveyed are very traditional in the capabilities they support — this is no doubt in

67 Making Archival and Special Collections More Accessible part an outcome of the type of functionality that from the Second World War. Filtering by a date is supported in EAD 2002. In addition to assessing range between 1939-1945 will result in only the suitability of the ArchiveGrid corpus for some those documents being presented that have a general archival-specific discovery interfaces, assigned in that period and may we wanted to cast our net a little wider and lead to the researcher missing potentially relevant speculate on how well EAD may meet the needs documents. Alternatively, only those documents of emerging NextGen (or NowGen!) approaches could be excluded that have a date outside to discovery that may not be represented in our of the indicated range. However, with a large interfaces surveyed, or supported by 2002 era amount of EADs missing a field EAD. Emerging discovery apparatus include: this approach defeats the purpose of filtering. Support for geo-locating archival locations, Investing effort to bring this element closer to subjects of collected materials, and high or complete may be warranted; however, other elements, to server map-based to support dimensions beyond just display, the search interfaces. Examples of map- content of the field or contents of the “normal” based discovery interfaces include: attribute must be easily parseable. When we look at the content of , we find • HistoryPin (http://www.historypin.com/), a wide range of descriptive practices, some of • WhatWasThere (http://www. which could pose problems for machine parsing whatwasthere.com/), to support use in indexing and retrieval.

• Historvius (http://www.historvius.com/) Another issue involved in using the Similarly, we see support for event-based field is that it can be used in several places, retrieval, using timelines or similar devices, e.g., on its own in the top level or as an area in which discovery systems are as a subelement of . evolving. Some examples include: Comparing the usage of in • SIMILE, example project timeline for Jewish our collection of EAD documents and that of History http://simile.mit.edu/timeline/ Wisser, we find that it is one of the elements examples/religions/jewish-history.html, where we see the greatest divergence, i.e., Wisser’s sample shows a usage of • Timeline view, Philippine Archives Collection, in the of 97.00%. NARA http://www.archives.gov/research/ military/ww2/philippine/timeline.html In ArchiveGrid, dates are used in: • Zagora Archaeological Project http://www. • search: they are keyword searchable powerhousemuseum.com/zagora/timeline/ • display: with the collection title (when available) in brief displays Analysis Details In other Archival Discovery Systems: • search We now take a closer look at which elements might drive each function, how the aggregated • browse data fits this purpose both in terms of meeting • sort our thresholds, and how well the content of key elements are fit for purpose. With each element, • display we’ve included a note about how they are used in ArchiveGrid and in other discovery systems. Extent Date Our analysis shows use of within the high-level as medium (70.43% — see Our analysis shows use of within Table 8); as with , the content the high-level as medium (72.64% — of is quite varied and does not see Table 7); This makes values easily facilitate sorting, with values ranging less than reliable for functions such as sort from “miscellaneous artifacts” to “2 ceramic and limit by. Consider, for example, a scenario heads.” The syntax of the element where a researcher is interested in material (with attributes for @encodinganalog, @ 68 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid type, and @unit) does not currently lend are fit to purpose. This would itself to structuring data in a way that can be be an opportunity for further evaluation, although used for sorting without clear guidelines, tools a quick scan of the contents of to enforce appropriate encoding, and rigor encouragingly revealed that 42% of ArchiveGrid on the part of institutions; retrospectively finding aids have a @type attribute with the refitting to be utilized in sorting could be a value “filing”, which is rather remarkable as daunting challenge for many institutions. there is no specified list of values for type. Many documents in the ArchiveGrid corpus In ArchiveGrid, collection titles are used in: have multiple statements, • search: they are keyword searchable further complicating matters, as the system would need to decide which one to sort, for • display: collection titles appear example. For display, including in brief search results statements in order to help aid researchers In other Archival Discovery Systems: in evaluating results seems fit to purpose. • sort In ArchiveGrid, extent is used in: • browse • search: extent values are keyword searchable • display • display: presented in brief displays and separately in the display of individual collection descriptions In other Archival Discovery Systems: • sort • display

Collection Title Our analysis shows use of in the high-level as complete (99.93% — see Table 7); this would suggest that it is suitable for all uses. However, for sorting and browsing, again, utility depends on the content of the element. If the content of the element is something generic like “Records” or “Papers” (in cases where perhaps the creator has been recorded separately in the origination element), then all functions may be less than ideal, but particularly sorting by title or creating browse lists. Many interfaces either construct browse lists of collections titles, or allow users to sort results by title, or search within titles. Not surprisingly, we found that the required element in the to be complete. Although our analysis did not include elements below , we can assume that the required and its required child, will be similarly complete. The fact that is fully populated is good news for searching and display; however for sorting and constructing browse lists, we would need to have some assurance that the contents of

69 Making Archival and Special Collections More Accessible

Subject Names (personal or corporate) Our analysis shows use of Names can be found in multiple places — for the as medium (72.89% — see Table 9); creator of a collection, is most logically found in is the parent element of , where both both subject as well as other access points (such and are child elements. The as , , , use of the origination tag is medium (87.78% and ). Our analysis did not include – see Table 7); our analysis did not include drilling down to use of evaluation of the use of and subelements. (Given differences in library and in origination. Otherwise, personal archival practices, we would expect control of and corporate names as access points may be form and genre terms to be relatively high, and found in (see above). control of names and subjects to be relatively low.) Name elements occur ubiquitously in EAD version In ArchiveGrid, subjects are used in: 2002, and our analysis did not include a detailed inventory of and • limit by: we show values for in the many places they can occur. A weakness of people, groups, places and topics as Result the distributed nature of names throughout EAD Overview facets for limiting a search result documents is that without detailed annotations In other Archival Discovery Systems: and co-references, discovery systems only • search have a shallow understanding of names and their relationship to the collection and to one • browse another. Discovery systems are not always able to differentiate between names when used in a creator context versus those covered in the Material type description, which may show up as access points. Researchers may wish to limit to or seek out In ArchiveGrid, names are used for: material in a specific format, and our survey of discovery systems reveal that some systems • search: names are keyword searchable support this functionality. Our analysis did not • limit by: names for people, groups and include the children of , places appear in the Result Overview which includes . In other Archival Discovery Systems: In ArchiveGrid, material type is used for: • Used in search • search: material types in are keyword searchable • Used for limiting In other Archival Discovery Systems: • search • browse • limit by

70 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

Repository Collections with digital content The name of the repository is found in the Our analysis did not explore the use of high-level did in . Use of this or elements, which can be used element falls into the promising complete in a variety of places in EAD 2002. Wisser category (99.46%: see Table 7). However, a and Dean found that is used in 7.7% variety of practice is in play, with the name and 9.3% of the documents in their sample, of the repository being embellished with putting both into the low category (see and

tags nested Wisser, Table 26, elements). However, within . To avoid the difficulties with growing interest in digitized materials in handling these variations across a range of from archival collections, identifying those contributing institutions, ArchiveGrid maintains materials is of increasing importance. a separate system to manage the form of the In ArchiveGrid, we provide no institution name for use in the system. mechanism for searching or identifying In ArchiveGrid, is not used collections with digital content. as an access point, though ArchiveGrid’s In other Archival Discovery Systems: separately administered and controlled form of the repository name is used for • Limiting results to those with digital content search, browse, sort, limit and display. • Flagging collections with digital content In other Archival Discovery Systems, used in: • browse Future Work • limit by In order to make EAD-encoded finding aids more well suited for use in discovery systems, Scope note, biographical the population of key elements will need to be moved closer to high or (ideally) complete. note, abstract However, it is not only a matter of populating the elements, but ensuring that the data will Our analysis shows use of as reliably power key aspects of discovery systems. high (84.41% — see Table 9), while This will take concerted effort and tools, both on (70.42% — see Table 9) and (79.20% — see Table 7) are medium; all three are the part of individual institutions and groups. suitable for search and for display in a results In the analysis of “NextGen” discovery services, view, although they can be quite lengthy. we noted the use of geolocation-based discovery. For search, its worth noting that the semantics Although we would need to do further analysis of these elements are different, and may result in to assess the usage for in unexpected and false “relevance” for matches in our document set, the current structure of the element does not against descriptions in (about the support geolocation functionality. However, as person) and and (which may be more about the collection). part of the redesign for EAD3, EAD is becoming more supportive of linked data and linked In ArchiveGrid, these notes are used in: data structures. This may offer some hope for • search: notes are keyword-searchable retrofitting EAD data to be more suited for the task of meeting map-based discovery requirements. • display: notes appear (in truncated form if Likewise, the data we have on hand does not lengthy) in brief search results suggest good support for event-based discovery, which would draw on well-structured dates, In other Archival Discovery Systems, used in: geographic subject terms, and topical subject • search terms (such as “Battle of Alma” or “”). Again, EAD 2002 does not support • display (in snippets or in their entirety) the sort of encoding that would be necessary to serve event-based discovery, but EAD3 may provide more appropriate structures.

71 Making Archival and Special Collections More Accessible

An Optimum Threshold So, is the container list half empty? If the archival community continues on its current path then for Discovery? the potential of the EAD format to support researchers or the public in discovery of material The picture for archival discovery and EAD is will remain underutilized. Minimally, collection decidedly mixed. On the one hand, we have descriptions that are below the thresholds for elements that are in high or even complete use. discovery will hinder their discovery efforts and On the other hand, we have many elements that maximally will remain hidden from view. Our are necessary for discovery interfaces that are paper provides suggestions for the elements in medium use; and even with elements that are where additional effort and investment are in high or complete use, the contents of those warranted to improve their utility for discovery tags are not always fit to purpose. This can be systems. (We recognize that for some institutions, at least partly explained by EAD’s history. In that additional effort may not be feasible or the early days of EAD the focus was largely on warranted; for their purposes they may find that moving finding aids from typescript to SGML and HTML or PDF collection descriptions suffice.) XML. Even with much attention given over to the development of institutional and consortial best Or is the container list half full? Perhaps with practice guidelines and requirements, much emerging evidence about the corpus of EAD, work was done by brute force and often with continued discussion of practice, recognition little attention given to (or funds allocated for) of a need for greater functionality, and shared making the data fit to the purpose of discovery. tools both to create new EAD documents and improve existing encoding, we can look Tag analyses such as the work described in forward to further increasing the effectiveness this paper can help inform the development and efficiency of EAD encoding, and develop a and implementation of the EAD schema practice of EAD encoding that pushes collection (indeed the work done by Wisser and Dean was descriptions across the threshold of discovery. considered in the development of EAD3). But our analysis suggests that the standard has most of the elements and attributes needed to effectively support discovery; what’s missing is agreement on and widespread application of best practices tied to supporting discovery.

72 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

Tables

Table 1: (Wisser Table 1): General statistics for EAD finding aids, using queries: /ead/*.

% % % [N_uniq/ Element N N_uniq [(N_uniqK)/ diff [N_uniq/S] n=124009] n=1136] eadheader 124009 124009 100.00 100.00 100.00 0.00 archdesc 124009 124009 100.00 100.00 100.00 0.00 frontmatter 46115 46115 37.19 37.19 24.60 12.59 eadgrp 0 0 0.00 0.00 0.00 0.00 archdescgrp 0 0 0.00 0.00 0.00 0.00 dscgrp 0 0 0.00 0.00 0.00 0.00

Table 2: (Wisser Table 2): Elements used within eadheader, using query /ead/eadheader/*.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1136] eadid 124445 124008 100.00 100.00 100.00 -0.00 filedesc 124009 124009 100.00 100.00 100.00 0.00 profiledesc 123103 123103 99.27 99.27 98.10 1.17 revisiondesc 42504 42501 34.27 34.27 32.70 1.57

Table 3: (Wisser Table 3) Attributes used with eadheader, using query //eadheader.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1136] countryencoding 107412 107412 86.62 86.62 89.50 -2.88 dateencoding 107377 107377 86.59 86.59 88.20 -1.61 findaidstatus 42910 42910 34.60 34.60 27.80 6.80

73 Making Archival and Special Collections More Accessible

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1136] langencoding 117641 117641 94.86 94.86 95.00 -0.14 repositoryencoding 106370 106370 85.78 85.78 87.80 -2.02 scriptencoding 95230 95230 76.79 76.79 77.60 -0.81

Table 4: (Wisser Table 4): Attributes used with eadid, using query //eadid.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1136] countrycode 108668 108667 87.63 87.63 94.30 -6.67 mainagencycode 105351 105350 84.95 84.95 92.60 -7.65 publicid 45758 45758 36.90 36.90 31.10 5.80 url 38020 38020 30.66 30.66 42.30 -11.64 urn 2312 2312 1.86 1.86 3.90 -2.04 identifier 57260 57260 46.17 46.17 49.30 -3.13

Table 5: (Wisser Table 8): Elements within frontmatter, using query /ead/frontmatter/*.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=46115] n=279] titlepage 45726 45726 36.87 99.16 92.80 6.36 div 190 190 0.15 0.41 2.20 -1.79

Table 6: (Wisser Table 9): Values for @level within archdesc, using query //archdesc/@level.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1,136] collection 116957 116957 94.31 94.31 90.90 3.41

74 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1,136] fonds 135 135 0.11 0.11 4.80 -4.69 class 9 9 0.01 0.01 0.30 -0.29 recordgrp 433 433 0.35 0.35 1.40 -1.05 series 2394 2394 1.93 1.93 0.60 1.33 subfonds 49 49 0.04 0.04 0.30 -0.26 subgrp 526 526 0.42 0.42 1.00 -0.58 subseries 46 46 0.04 0.04 0.00 0.04 file 2446 2446 1.97 1.97 0.40 1.57 item 987 987 0.80 0.80 0.30 0.50 otherlevel 25 25 0.02 0.02 0.10 -0.08

Table 7: (Wisser Table 10): Elements within archdesc/did, using query /ead/archdesc/did/*.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1,136] abstract 102792 98218 79.20 79.20 86.60 -7.40 container 5447 3471 2.80 2.80 0.40 2.40 langmaterial 112938 109232 88.08 88.08 89.90 -1.82 materialspec 41 41 0.03 0.03 1.60 -1.57 origination 113684 108853 87.78 87.78 89.00 -1.22 physdesc 135126 122402 98.70 98.70 97.20 1.50 physloc 53564 45620 36.79 36.79 27.80 8.99 repository 123343 123330 99.45 99.45 99.60 -0.15 unitdate 97247 90080 72.64 72.64 97.00 -24.36 unitid 119911 114898 92.65 92.65 90.10 2.55

75 Making Archival and Special Collections More Accessible

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1,136] unittitle 123959 123916 99.93 99.93 100.00 -0.07

Table 8: (Wisser Table 11): Elements within archdesc/did/ physdesc, using query /ead/archdesc/did/physdesc/*.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1,136] dimensions 666 576 0.46 0.46 1.80 -1.34 extent 122613 87339 70.43 70.43 76.30 -5.87 physfacet 2000 1520 1.23 1.23 1.70 -0.47

Table 9: (Wisser Table 12): Elements within archdesc:above the dsc, using query /ead/archdesc/*.

% [(N_ % [N_ % [N_uniq/ Element N N_uniq uniqK)/ diff uniq/S] n=124009] n=1,136] accessrestrict 55751 55579 44.82 44.82 86.20 -41.38 accruals 694 694 0.56 0.56 7.10 -6.54 acqinfo 40668 40451 32.62 32.62 68.00 -35.38 altformavail 2293 2289 1.85 1.85 12.70 -10.85 appraisal 4613 4602 3.71 3.71 4.80 -1.09 arrangement 40979 40627 32.76 32.76 65.50 -32.74 4573 4083 3.29 3.29 10.10 -6.81 bioghist 89103 87333 70.42 70.42 87.30 -16.88 controlaccess 92124 90390 72.89 72.89 85.00 -12.11 custodhist 8375 8366 6.75 6.75 14.10 -7.35 descgrp 67684 56446 45.52 45.52 32.00 13.52 fileplan 50 44 0.04 0.04 0.60 -0.56

76 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

% [(N_ % [N_ % [N_uniq/ Element N N_uniq uniqK)/ diff uniq/S] n=124009] n=1,136] index 1231 656 0.53 0.53 1.20 -0.67 odd 9594 8145 6.57 6.57 9.70 -3.13 originalsloc 988 973 0.78 0.78 3.40 -2.62 otherfindaid 6529 6271 5.06 5.06 11.90 -6.84 phystech 900 897 0.72 0.72 4.20 -3.48 prefercite 49015 48989 39.50 39.50 85.40 -45.90 processinfo 27249 26623 21.47 21.47 0.00 21.47 relatedmaterial 23932 23676 19.09 19.09 40.30 -21.21 runner 10822 10822 8.73 8.73 1.10 7.63 scopecontent 105384 104670 84.41 84.41 93.40 -8.99 separatedmaterial 5789 5691 4.59 4.59 14.80 -10.21 userestrict 41365 40749 32.86 32.86 68.30 -35.44

Table 10: Table 13: The inclusion of dsc in finding aids, using query //dsc.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=124009] n=1,136] < dsc > 98663 94473 76.18 76.18 90.30 -14.12 multiple 98663 2075 1.67 1.67 2.40 -0.73 < dsc > s

Table 11: (Wisser Table 14): dsc type attributes, using query //dsc/@type.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=99023] n=1,105] analyticover 3156 3149 2.54 3.18 5.10 -1.92 combined 49205 49184 39.66 49.67 66.50 -16.83

77 Making Archival and Special Collections More Accessible

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=99023] n=1,105] in-depth 36433 35876 28.93 36.23 16.70 19.53 othertype 1725 1572 1.27 1.59 3.50 -1.91

Table 12: (Wisser Table 15): c-c12 tags, using query //c | //c01 | //c02 | // c03 | //c04 | //c05 | //c06 | //c07 | //c08 | //c09 | //c10 | //c11 | //c12.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=96548] n=1,053] c 4745698 14440 11.64 14.96 11.10 3.86 c01 1650659 78600 63.38 81.41 88.00 -6.59 c02 7432993 59217 47.75 61.33 72.50 -11.17 c03 6625963 29136 23.50 30.18 41.80 -11.62 c04 2927180 12819 10.34 13.28 20.60 -7.32 c05 1312217 5587 4.51 5.79 10.70 -4.91 c06 598647 2266 1.83 2.35 4.60 -2.25 c07 261648 922 0.74 0.95 2.00 -1.05 c08 90401 331 0.27 0.34 0.70 -0.36 c09 21514 110 0.09 0.11 0.30 -0.19 c10 3578 36 0.03 0.04 0.10 -0.06 c11 823 7 0.01 0.01 0.00 0.01 c12 96 2 0.00 0.00 0.00 0.00

78 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

Table 13: (Wisser Table 16): Values for level attribute on c, c/@level, using query //c/@ level | //c01/@level | //c02/@level | //c03/@level | //c04/@level | //c05/@level | //c06/@level | //c07/@level | //c08/@level | //c09/@level | //c10/@level | //c11/@level | //c12/@level.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=96548] n=1,053] collection 13489 4782 3.86 4.95 2.10 2.85 fonds 418 95 0.08 0.10 0.70 -0.60 class 63134 2113 1.70 2.19 1.20 0.99 recordgrp 1535 193 0.16 0.20 0.70 -0.50 series 398727 58480 47.16 60.57 77.70 -17.13 subfonds 3210 637 0.51 0.66 1.70 -1.04 subgrp 5573 430 0.35 0.45 3.10 -2.65 subseries 466366 16974 13.69 17.58 35.30 -17.72 file 11419524 36262 29.24 37.56 56.90 -19.34 item 3480272 20415 16.46 21.14 24.20 -3.06 otherlevel 368942 6225 5.02 6.45 9.10 -2.65

Table 14: (Wisser Table 17): c-c12/did elements, using query //c/did/* | //c01/ did/* | //c02/did/* | //c03/did/* | //c04/did/* | //c05/did/* | //c06/did/* | //c07/ did/* | //c08/did/* | //c09/did/* | //c10/did/* | //c11/did/* | //c12/did/*.

% % [N_uniq/ Element N N_uniq % [N_uniq/S] [(N_uniqK)/ diff n=96548] n=1,053] abstract 1421043 3850 3.10 3.99 2.50 1.49 container 24951558 72377 58.36 74.96 82.50 -7.54 langmaterial 46798 1127 0.91 1.17 6.10 -4.93 materialspec 22870 106 0.09 0.11 1.30 -1.19 origination 1308346 4090 3.30 4.24 8.10 -3.86 physdesc 3967094 37749 30.44 39.10 54.40 -15.30

79 Making Archival and Special Collections More Accessible

% % [N_uniq/ Element N N_uniq % [N_uniq/S] [(N_uniqK)/ diff n=96548] n=1,053] physloc 1343791 5978 4.82 6.19 5.80 0.39 repository 34923 29 0.02 0.03 0.30 -0.27 unitdate 9613593 41894 33.78 43.39 90.60 -47.21 unitid 7167784 31035 25.03 32.14 46.20 -14.06 unittitle 25228059 92888 74.90 96.21 98.90 -2.69

Table 15: (Wisser Table 18): c-c12/did/physcdesc elements, using query //c/did/physdesc/* | // c01/did/physdesc/* | //c02/did/physdesc/* | //c03/did/physdesc/* | //c04/did/physdesc/* | // c05/did/physdesc/* | //c06/did/physdesc/* | //c07/did/physdesc/* | //c08/did/physdesc/* | // c09/did/physdesc/* | //c10/did/physdesc/* | //c11/did/physdesc/* | //c12/did/physdesc/*.

% % [N_uniq/ Element N N_uniq % [N_uniq/S] [(N_uniqK)/ diff n=96548] n=1,053] dimensions 144079 1378 1.11 1.43 5.20 -3.77 extent 2401903 24495 19.75 25.37 36.60 -11.23 physfacet 164430 613 0.49 0.63 6.80 -6.17

Table 16: (Wisser Table 19): other elements found in c-c12, using query //c/* | //c01/* | // c02/* | //c03/* | //c04/* | //c05/* | //c06/* | //c07/* | //c08/* | //c09/* | //c10/* | //c11/* | //c12/*.

% % [N_ % [N_uniq/ Element N N_uniq [(N_uniqK)/ diff uniq/S] n=96548] n=1,053] accessrestrict 600069 4844 3.91 5.02 10.70 -5.68 accruals 12 11 0.01 0.01 0.00 0.01 acqinfo 68066 1477 1.19 1.53 4.50 -2.97 altformavail 252282 766 0.62 0.79 2.70 -1.91 appraisal 48 30 0.02 0.03 0.70 -0.67 arrangement 31945 5746 4.63 5.95 19.00 -13.05

80 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

% % [N_ % [N_uniq/ Element N N_uniq [(N_uniqK)/ diff uniq/S] n=96548] n=1,053] bibliography 2067 48 0.04 0.05 1.50 -1.45 bioghist 12511 1132 0.91 1.17 4.60 -3.43 controlaccess 243134 2149 1.73 2.23 5.10 -2.87 custodhist 26224 181 0.15 0.19 2.20 -2.01 descgrp 2703 31 0.02 0.03 1.80 -1.77 index 386148 835 0.67 0.86 0.70 0.16 note 1180397 11265 9.08 11.67 20.30 -8.63 odd 242182 2663 2.15 2.76 7.20 -4.44 originalsloc 9959 211 0.17 0.22 1.00 -0.78 otherfindaid 1945 247 0.20 0.26 2.30 -2.04 phystech 8439 300 0.24 0.31 1.50 -1.19 prefercite 1995 264 0.21 0.27 0.10 0.17 processinfo 26332 1084 0.87 1.12 3.80 -2.68 relatedmaterial 16727 882 0.71 0.91 4.40 -3.49 runner 0 0 0.00 0.00 0.00 0.00 scopecontent 1852092 33483 27.00 34.68 61.30 -26.62 separatedmaterial 2784 208 0.17 0.22 0.00 0.22 userestrict 2993 580 0.47 0.60 3.20 -2.60

Table 17: (Wisser Table 20): content tags in dsc, using query //dsc//*.

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=96548] n=1,053] corpname 373402 6082 4.90 6.30 8.40 -2.10 famname 3644 914 0.74 0.95 1.70 -0.75

81 Making Archival and Special Collections More Accessible

% [(N_ % [N_uniq/ Element N N_uniq % [N_uniq/S] uniqK)/ diff n=96548] n=1,053] function 996 53 0.04 0.05 0.00 0.05 genreform 351956 6988 5.64 7.24 5.10 2.14 geogname 1023771 6653 5.36 6.89 6.30 0.59 name 34339 380 0.31 0.39 1.40 -1.01 occupation 25284 285 0.23 0.30 0.40 -0.10 persname 2610548 11970 9.65 12.40 12.90 -0.50 subject 1239139 2419 1.95 2.51 4.70 -2.19

82 Thresholds for Discovery: EAD Tag Analysis in ArchiveGrid

References About the Authors

1 In April 2013, the ArchiveGrid index Marc Bron is a researcher at the Intelligent contained 1,632,246 MARC records, Systems Lab , where he is about to 119,984 EAD records, 61,551 HTML complete his PhD in Information Retrieval. His records, and 4,532 PDF records. The EAD dissertation focused on improving accessibility to count in the index is lower than the set of information stored in cultural heritage institutions documents analyzed, to avoid duplicating by developing algorithms and interactive their display for certain contributors who retrieval systems that support exploration and supply corresponding MARC records. contextualization. During his PhD Marc has published over 20 papers at top tier conferences, 2 Library of Congress EAD Website: journals, and workshops. His current research http://www.loc.gov/ead/index. direction aims to develop new collaborative html; EADiva: http://eadiva.com/. search methods for users of archival collections. 3 E-mail to archives and archivists Bruce Washburn is a Consulting Software listserv, November 15, 2010. Engineer in OCLC Research. He provides software 4 Wisser, Katherine M, and Jackie Dean, development support for OCLC Research EAD Tag Usage: Community analysis of initiatives and participates as a contributing the use of Encoded Archival Description team member on selected research projects. elements, article submitted for In addition, he provides software development publication in the American Archivist support for selected OCLC Products and Services. At OCLC Washburn has been a part of the product 5 Smith-Yoshimura, Karen, Catherine Argus, Timothy J. Dickey, Chew Chiat teams that developed and maintain CAMIO, Naun, Lisa Rowlinson de Ortiz, and Hugh ArchiveGrid, the WorldCat Search API, and OAIster. Taylor. 2010. Implications of MARC Tag Merrilee Proffitt is a Senior Program Officer Usage on Library Metadata Practices. in OCLC Research. She provides skills and expert support to institutions represented within the OCLC Research Library Partnership. Merrilee has authored or co-authored articles, guidelines, and reports for a variety organizations and professional journals. She is frequently an invited speaker at international professional conferences and workshops on topics relating to digital libraries and special collections. Her current projects and interests include: archival description, increasing access to special collections, looking at developing better relationships between and cultural heritage institutions, and how Massively Open Online Courseware (MOOCs) may impact libraries. She is a member of the small but mighty ArchiveGrid team.

83 Making Archival and Special Collections More Accessible

84 Making Archival and Special Collections More Accessible 7 The Metadata is the Interface: Better Description for Better Discovery of Archives and Special Collections,Synthesized from User Studies

Jennifer Schaffner Program Officer OCLC Research

This paper was originally published in May 2009 by OCLC Research at http://www.oclc.org/content/dam/research/publications/library/2009/2009-06.pdf.

85 Making Archival and Special Collections More Accessible

Introduction

Tim Ericson warned that user studies are always been to improve practices in order to important, but “they can also be a substitute for help people—not just archivists and librarians— more direct action.”1 We have strong evidence discover archival and rare materials.10 We still about how to improve discovery of archives have gaps in our understanding, and comparing and special collections, and we need to start different kinds of studies across many years somewhere. These days we are writing finding of work is like comparing apples and oranges. aids and cataloging collections largely to be Nevertheless, the community has learned from discovered by search engines. People expect to these studies about obstacles between people find archives and special collections on the open and unique materials. While there is more to Web using the same techniques they use to find learn, let’s start now by adjusting our practices other things, and they expect comprehensive in order to disclose information about special results. Invisibility of archives, manuscripts and collections and archives more effectively. special collections may well have more to do with the metadata we create than with the interfaces we build. Now that we no longer control discovery, Librarians and Archivists the metadata that we contribute is critical. In as Gatekeepers so many ways, the metadata is the interface.2 Users work increasingly on their own, while Structured metadata can be useful internally librarians and archivists have expected to for collection management and public services, mediate research. Most often people want to be but is not always what users need most to autonomous and discover information about discover primary sources, especially minimally- primary sources at the network level, not at 3 described collections and “hidden collections.” the institutional level.11 In an Ithaka study of We understand archival standards for description higher education, Roger Schoenfeld and Ross 4 and cataloging, but our users by and large don’t. Housewright learned that scholars consider less Studies show that users often do not want to mediation in research and discovery a good thing: search for collections by provenance, for example, as important as this principle is for archival [L]eading-edge libraries are beginning to collections.5 One of several core competencies change their priorities to match those of that special collections metadata librarians must faculty and students. Still, the mismatch in have is “a keen understanding of users’ needs and views on the gateway function is a cause preferences.”6 This is especially important now for further reflection: if librarians view that discovery happens in multiple environments.7 this function as critical, but faculty in Librarians and archivists need to manage archival certain disciplines find it to be declining in collections by provenance, but also must describe importance, how can libraries, individually or what is in the collections for their users. collectively, strategically realign the services that support the gateway function?”12 This essay—part of a series of OCLC Research projects to mobilize unique materials—synthesizes The more that discovery occurs directly via search evidence of what descriptive information people engines, the greater the success of considerable say they need for research.8 As this literature efforts to expose “hidden collections.” review got underway, it soon became evident Over twenty-five years ago, Mary Jo Pugh that we already know most of what we need to challenged the myth of immortal and omniscient know in order to get started making changes. archivists, on whom users would rely for access 13 In many contexts over many years, librarians to the contents of archival collections. Many and archivists have studied users with a wide studies of library catalogs and archival portals variety of research methods: using surveys have shown that these days most users start and questionnaires, examining statistics and their search for information with Google or citations, testing usability of interfaces, studying Wikipedia, and usually only come to libraries 14 information-seeking behaviors, listening to focus and archives for known items. Now the primary groups, creating personas, and questioning the role in discovery is making the collections efficacy of finding aid portals.9 The goal has more visible and staying out of the way: 86 The Metadata is the Interface: Better Description for Better Discovery

“Perceptions of a decline in dependence Content is more important than format.25 Over are probably unavoidable as services are fifteen years ago, Jackie Dooley cautioned that increasingly provided remotely, and in without subject access to records about archival some ways these shifting faculty attitudes collections, users are reduced to known-item can be viewed as a sign of library success. searching.26 An example of this surfaced in One can argue that the library is serving recent usability testing of WorldCat Local at the faculty well, providing them with a less University of California (UC). Faculty and graduate mediated research workflow and greater student participants only searched UC’s union ability to perform their work more quickly catalog for known items, not for discovery, when and effectively. In the process, however, they they were working in their areas of expertise.27 In may be making their own role less visible.”15 an example of good intentions, the Online Archive of California (OAC) hoped to add subject searching Perhaps ironically, goals to disclose until they learned that only “60% of the finding descriptions online and to digitize primary aids used controlled access tags.”28 Richard Szary resources have made special collections more and Lawrence Dowler recommended “direct visible and roles of archivists and librarians indexing of the content of historical materials” less visible. The more users do not need to to improve access.29 For discovery, Aboutness consult archivists and librarians for searching, is a very important element of description. the more successful initiatives to improve description and discovery have been. While users want to find subjects, they generally search using keyword techniques, rather than by using structured terminology. For example, Users Search by Subjects research shows that keywords are important to and Keywords historians searching for known items.30 Likewise, NWDA usability testers observed that searches Archivists and librarians have often focused on were completely unstructured.31 In November what collections are made up of (Ofness), while 2008, the French CALAMES project reported many users prefer to learn what collections are 40% frequency of searching full text, 34% by 16 about (Aboutness). Studies report consistently personal name, and 19% by various subject that many users want to find information about elements and attributes.32 Susan Hamburger’s 17 contents of collections. For instance, Bill Maher research yielded different proportions: 78% analyzed reference letters to the University of by keywords, 31% by names and 23% by Illinois archives in 1984-85 and found that over subjects.33 Chris Prom also found that users one third of the researchers inquired about of the University of Illinois’s electronic finding 18 subjects. One respondent in Jane Stevenson’s aids primarily used non-fielded keyword search testing of the UK’s Archives Hub said, “I like the terms, along with structured browsing.34 subject finder. I’m pleasantly surprised by it.”19 In the most recent Northwest Digital Archives Recent work addressed phrase-searching (NWDA) usability test, one user was enthusiastic techniques. Phrase searches have been shown to discover the subject section: “These will give to be more effective than keyword searches 35 me an idea of what this collection is about.”20 when using search engines to find finding aids. In a previous NWDA usability study, one person People don’t search that way, however, according recommended controlled subject vocabulary and many studies, including OAC usability testing in 36 wanted subject terms linked to other collections 2001. In another example, 8 out of 9 participants and catalogs.21 Wendy Duff concluded, in more searched by keywords—not phrases—in NWDA 37 than one study, that users wanted “what is usability testing. Kristina Southwell used it about?” to appear at first glance.22 Louise statistical reports from search engines to Gagnon-Arguin found 41% of queries in Québecois demonstrate that the University of Oklahoma’s archives were for subjects or themes.23 A study Web pages for manuscripts were typically in 1976 of registration forms at the Michigan found through keyword searches, although 38 Historical Society showed that, “Roughly half of some people used subject phrases, too. Based all users, regardless of preparation, began with on research with users at six major research a subject searching approach.”24 For thirty years, libraries, Susan Hamburger recommended people have reported that they want to discover offering searching on both keywords and subject 39 archival materials using subject information. terms in catalog records and finding aids.

87 Making Archival and Special Collections More Accessible

A wide range of research shows that keyword discovery experiences that are simple, elegant searching is important specifically for humanities and intuitive.49 Users want to search names by scholars, who often search using name, place, keyword, search for subjects by browsing, and title and discipline-specific terms.40 Jihyun Kim browse by keyword or name, too. When it comes examined EAD finding aids themselves, rather to using descriptive metadata to discover archival than users, precisely because historians and materials and special collections, users want humanists search for primary sources by names it all. This is problematic because significant of people and places. Kim reported that few principles of archival theory and practice have finding aids used “controlled access headings.”41 been provenance and description of what the Wendy Duff and Catherine Johnson interviewed collection is made up of, its Ofness.50 For users, ten historians and concluded they search names research shows that important elements of primarily because names are the easiest way description, especially minimum-level description, into collections. Social historians desired subject are keywords and terms that indicate Aboutness. access to collections: “’There has to be a way that people can find things without having to know who generates them, so keywords will Users Expect Results search across different of things’ Ranked by Relevance (participant 6).”42 Using keyword searching techniques for topics— such as farm women— While researchers consider it important to know can be problematic, because archives are the relative importance of collections, archivists organized primarily by the names of the creators, and librarians rarely create metadata that can be not the subject content of the collections. used to rank relevance. In 1987, Avra Michelson argued that scholars using primary sources There is no common understanding of what expected relevant results when doing research users and testers mean when they use words like in exhaustive listings of collections.51 Over “keyword,” “subject,” “known item,” “name,” twenty years later, students at the University of “phrase” and “browse.” Without that common Maryland were overwhelmed by large result sets understanding, it is difficult to compare findings retrieved by keyword searches; they expected from separate studies. Is a keyword search relevance ranking of results such as that returned technique in effect a subject search, from a user’s by Google and other search engines.52 Chris 43 standpoint? In one test, while Archives Hub Prom learned— using the interface for the participants favored subject searching, they were University of Illinois Archives—that hits sorted by confused by a browse list composed of access provenance confused his participants, who were 44 points. Do testers consider natural language largely expecting search results to be ranked by searching to be keyword or subject searching, relevance.53 Andrea Rosenbusch concluded, after even if the user’s search includes names? Wendy studying a dozen archival online databases, that, Duff and Catherine Johnson, for example, consider “The relevance of provenance as the main access a search by name keywords to be a known-item point to records is becoming questionable…”54 search.45 Users do not always distinguish clearly between names and subjects. For instance, two As it stands now, identifying relevant primary of the participants in the Archives of American Art resources often requires educated guesswork. usability study never found the Joseph Cornell All of the participants in Sara Snyder’s study collection because they searched by keywords at the Archives of American Art said that rather than browsing an alphabetical list of relevance ranking was essential, especially 55 collections.46 RLG learned from focus groups that for large results. On the other hand, in Jane many participants combine keywords with names, Stevenson’s Archives Hub study, relevance subjects and dates.47 Richard Lytle speculated ranking of the results of a subject search puzzled that many kinds of searches might be disguised some people, who then wanted to know how subject searches: “Requests for records by proper relevance worked and why some hits were 56 name, geographical area, date or form may more relevant. When redesigning ArchiveGrid conceal a subject request. Does the user really for improved usability, RLG determined that prefer to ask for documents by name…?”48 the order of search results was important, and relevance—not title—was the desired order.57 Elsie Freeman memorably posited that good subject information is a large component of

88 The Metadata is the Interface: Better Description for Better Discovery

Several tactics have been proposed that could Comprehensive Coverage indicate the relative importance of special collections in discovery experiences. Extent or Increasingly, archivists and librarians are physical description elements can be useful for acutely aware that many researchers expect some researchers trying to sort out relevance comprehensive coverage. A student in the for themselves. (“Just one quick question. Does Maryland study expected that “the universe anybody understand what twelve metres of of primary sources is a finite, absolute body of textual records means?” “Means two weeks in material that can and has been already labeled 67 the archives!”58) Andrea Rosenbusch suggested and categorized for him.” Chris Prom, too, relevance could be leveraged from multi-level learned that many inexperienced users assume 68 description, by restricting queries to top-level that everything is available. Jane Stevenson’s descriptions: “The aim [of ISAD(G)] is to enable study with Archives Hub confirmed that some users to identify fonds or even whole collections people assumed their search results were 69 which have the highest relevancy to them.”59 comprehensive. In a usability study of the Systems don’t exist yet that use standards-based Lilly Library’s Web site, Erika Dowell found that descriptions and extent statements in this way. users doubted the utility of the online catalog when cautioned (responsibly) that only 45% of Search engine optimization strategies could the Lilly’s holdings were included.70 In a related leverage metadata for sorting search results by study in UK museums, the Research Information relevance. Based on keyword density analysis of Network concluded that “what researchers UC Irvine’s finding aids, Michelle Light advocated need above all is online access to the records in enhancing discovery by describing collections museum and collection databases to be provided more strategically—by using more keywords and as quickly as possible, whatever the perceived 60 concepts than folder lists and material types. imperfections or gaps in the records.”71 Taking another tactic, the NWDA Working Group recommended experimenting with algorithms Some researchers have substantiated a “More to combine use statistics with the frequency Product, Less Processing” (MPLP) approach to 72 of index terms in order to produce relevance description and digitization. At the University ranking like in search engines.61 Recommender of Wisconsin, Joshua Ranger and Krystyna systems for discovery of archival collections Matusiak are experimenting with a less expensive, might provide indications of relevance. streamlined process for mass digitization Improvements will require imaginative use of of archival collections. The students they available Web 2.0 tools, such as tags for important interviewed all preferred more description, not collections on a topic, or “link paths” like those less. However, when the comparative costs of demonstrated in the Polar Bear project.62 full and minimal records were explained, all of the participants said streamlined description Over twenty years ago, Avra Michelson called for was preferred: “Better than not having it at all.”73 study of search questions, in order to identify The American Heritage Center at the University successful patterns. Michelson recommended of Wyoming surveyed 600 respondents for their subsequent improvements in our use of subject satisfaction with minimal processing. Asked to terms in description in order to improve what she rank archival priorities, respondents most often 63 called “retrieval capabilities.” More recently, chose “putting more resources into creating Karen Markey has similarly suggested we would basic descriptions for all collections.”74 The learn a great deal from studying people’s search MPLP approach matches users’ acceptance 64 terms. Many user studies for archives and of minimum-level description because they special collections have focused on discovery would like to discover more materials online. within local systems designed for archival When such decisions are made to describe more 65 materials. Now that close to 90% of searching collections at a minimal level, archivists and 66 behavior begins in search engines, it is time to librarians need to indicate, however briefly, evaluate search behaviors at the network level, what the contents of collections are about. in order to develop descriptive strategies for ranking the relevance of primary resources.

89 Making Archival and Special Collections More Accessible

Users Know How to Users’ Lack of Awareness Scan and Scroll Archivists and librarians have created catalogs and portals, but many users don’t use them or Archivists and librarians have worried about don’t know they exist. “The greatest barrier to confusing users by presenting different kinds use is lack of awareness.”82 Often it isn’t easy and amounts of metadata, while users mostly to find rare and unique library and archival care more about what is in the collections.75 materials because successful discovery currently Diverging desires for less or more information requires people to understand what they are appear often in user research. One example of looking for and how to find it.83 Karen Markey this variety is that Maryland students were able says rare and unique materials are invisible: to use long finding aids easily, despite difficulties “Thousands of special collections that make navigating specific tools to discover them. On up the invisible Web feature their own unique the other hand, in the same test one Maryland search engines because their content is not student reported that “too much information accessible via general Web search engines.”84 hindered the reading of the display.”76 In another Louise Gagnon-Arguin concluded that the key contradictory example, some of Wendy Duff’s to access is fragile in the context of electronic participants preferred to see shorter abstracts information.85 In order to find primary resources, and scope-and-content notes, disparaging long people need to know too much about how biographical notes (ranked 16th in order of collections are described and where those preference) or administrative histories (ranked descriptions are lodged. That isn’t good enough. 23rd). However, a different participant in the same study said anyone interested in long Catalogs don’t seem to do the trick. “It is unlikely notes can scroll down through the display.77 that researchers approach doing research by looking for a tool for doing research.”86 In the RIN There are many more examples of preferences user study of UK museums, “most researchers for both brief and for full displays that support are unaware of the online catalogues…”87 Beth arguments for both minimal and full description. Yakel, Susan Hamburger, Bill Maher and others In the RLG rapid iterative interface testing, most have found that the majority of researchers do participants found a brief scope-and-content note not use utilities such as ArchivesUSA, OCLC, most useful.78 On the other hand, studies also RLIN or NUCMC.88 While a percentage of people report that users know how to skim long pages of in Kristina Southwell’s Oklahoma survey found records, when they want to. In Jane Lee’s usability manuscript collections by searching the Web, testing for the 2008 OAC redesign, she noted that only one person (0.4% in 230 responses!) used they chose a long display format for search results RLIN’s AMC.89 Southwell was surprised that only because, as one participant said, “it’s nice to have 11.3% of respondents discovered manuscript a little more information” when browsing.79 In collections using the online catalog, leading her NWDA usability testing, “the majority of the users to wonder about the considerable investment started the search for information by skimming creating MARC records. 17.9% found collections or scrolling through the finding aid page; most from html finding aids on the Web site, 25.1% said they weren’t reading for content, rather were used footnotes and , while another scanning for key terms.”80 Genealogists in RLG’s 8.6% used a published guide to the repository for Archival Resources focus groups preferred to know-item discovery. Users may search on the scroll through large result sets.81 These conflicting open Web, but often they find archives indirectly. recommendations suggest that minimum description may come as a relief to some users, So are finding aids best for discovery? Bill Maher but others prefer a full description. If a collection questioned out loud our tacit belief that better is fortunate enough to have full description, it will finding aids will automatically result in better not necessarily get in a user’s way when scanning access.90 Most participants in Jane Stevenson’s and scrolling through results. Users support study of the Archives Hub “did not mention any concise minimum-level description, which can kind of cross-searching networks.”91 Kathleen also be effective for discovery when it is done well. Feeney concluded that “electronic finding aids may not be well suited to serve as pointers to archival collections,” based on her 1999 study of retrieval of full-text finding aids by search

90 The Metadata is the Interface: Better Description for Better Discovery engines. Feeney concluded that “MARC records Conclusion remain a more valuable and reliable means of locating archival resources” because of problems I argue that some thirty years of user studies with relevance rankings at the network level.92 teach that Aboutness and relevance matter most for discovery of special collections, especially For successful discovery, what are the lessons now that discovery happens elsewhere.95 learned about our choices for description? Early Unfortunately, there is a gap between the on, Rob Spindler and Richard Pearce- expectations of users and historical descriptive argued for adapting description methods— practices in archives and special collections. based on their case study with Arizona State Changes must be made to description because University patrons—expressly to improve researchers rarely look in library catalogs comprehension of AMC records in an integrated 93 or archival portals for primary resources. online environment. More recently, Michelle These changes are even more important Light suggested strategies to adapt description for collections that have been selected for than can “enhance retrieval possibilities” at minimal processing and description. Ensuring the network level: use long-tail keywords, that “hidden collections” can be discovered repeat names and keywords (bending rules requires appropriate description, not just for description), put the most important 94 expert processing, cataloging and cross- content at the top, say more with less. If searching networks. It would be heartbreaking students now don’t look in library catalogs if special collections and archives remained or archival portals for primary materials, why invisible because they might not have spend resources that way? Let’s put the right the kinds of metadata that can easily be descriptive metadata in the right places. discovered by users on the open Web. In a 1986 article on “The Use of User Studies,” Bill Maher described archivists with instincts about how their collections are used—but without data to support their instincts—as “working in the dark.”96 Since then, research demonstrates recurring observations of users’ needs and preferences when they search for special collections and archives. Over time, users have adapted their research tactics: from discovery only by visiting repositories and by consulting printed catalogs or guides, then discovery using online catalogs and portals, and now discovery on the Web. All along, user studies have demonstrated that descriptive metadata indicating Aboutness and relevance matters significantly for discovery. Twenty years later, we are not working in the dark any more.

91 Making Archival and Special Collections More Accessible

Notes 8 I focus attention on metadata for discovery, not on research methodologies, 1 Timothy L. Ericson, “Preoccupied with Our nor on interfaces, nor on interface Own Gardens: Outreach and Archives,” terminology, nor on information- Archivaria 31 (winter 1990-1991): 120. seeking behavior writ large, nor on Richard Cox expressed similar concerns: what users do with what they find, “User studies and citation analyses float much as these topics are fascinating isolated from any practical applications.” and closely related to the relationship See “Researching Archival Reference as between description and discovery. an Information Function: Observations on Needs and Opportunities,” RQ 9 Many more studies have been done 31, no.3 (spring 1992): 393. than are reported: “In recent years archive services in many parts of the 2 This aphorism was coined world have undertaken surveys of by Arnold Arcolio. their users. Valuable work has been done, but most of it suffers from lack 3 Judith M. Panitch conducted the first survey that catalyzed the community of availability–many surveys remain around the significance of “hidden unpublished and largely inaccessible…” collections”: Special Collections in ARL Geoffrey Yeo, “Understanding Users Libraries (Washington DC: Association and Use: A Market Segmentation of Research Libraries, 2001). Published Approach,” Journal of the Society of online at: http://www.arl.org/bm~doc/ Archivists 26, no. 1 (April 2005): 43-44. spec_colls_in_arl.pdf. See also Barbara 10 Lawrence Dowler called for a national M. Jones, “Hidden Collections, Scholarly study of the use and users of archives Barriers” (June 6, 2003). Published to inform archival practices: “The Role online at: http://www.arl.org/bm~doc/ of Use in Defining Archival Practice and hiddencollswhitepaperjun6.pdf. Principles,” American Archivist 51 (winter/ spring 1988): 74-95. A stellar example is 4 I have borrowed this idea from Steve Hensen. the National Council on Archives in the UK, which established standards requiring 5 Wendy M. Duff and Catherine A. Johnson, repositories to give users a choice in “Accidentally Found on Purpose: access to content, and to open up that Information-Seeking Behavior of content to a wider range of users: David Historians in Archives,” Library Quarterly Mander, ed., Standard for Access to Archives 72, no. 4 (2002): 477, footnote 5. (National Council on Archives [UK]: Public nd 6 ACRL Rare Books and Manuscripts Services Quality Group, 1st ed. 1999, 2 rd Section Task Force on Core Competencies ed. 2003, 3 ed. 2008), 37, 46. The Auditor for Special Collections Professionals, General of Canada is taking a similar “RBMS Competencies for Special tactic: Jean Dryden, “Do We Care What Collections Professionals,” College Users Want? Evaluating User Satisfaction and Research Libraries News, 69, and the LibQUAL+™ Experience,” Journal of no. 10 (November 2008): 622. Archival Organization 4, no. 4 (2004): 83-84.

7 Lorcan Dempsey, “The Library Catalogue 11 Louise Gagnon-Arguin applied the in the New Discovery Environment: term “autonomy” to users doing Some Thoughts,” Ariadne 48 (July research: “Les questions de recherché 2006). Published online at: http://www. comme matériau d’études des usagers ariadne.ac.uk/issue48/dempsey/. en vue du traitement des archives,” Archivaria 46 (winter 1998): 86-102.

92 The Metadata is the Interface: Better Description for Better Discovery

12 Ross Housewright and Roger Schoenfeld, 17 For example, a recent study of users of Ithaka’s 2006 Studies of Key Stakeholders online library catalogs found that users in the Digital Transformation in Higher desired “more subject information” Education (August 2008). Published online second only to more online content: at: http://www.ithaka.org/research/ Karen Calhoun, Online Catalogs: What Ithakas%202006%20Studies%20 Users and Librarians Want (Dublin, of%20Key%20Stakeholders%20in%20 Ohio: OCLC, 2009), 13, 43, 47. Published the%20Digital%20Transformation%20 online at: http://www.oclc.org/reports/ in%20Higher%20Education.pdf. onlinecatalogs/fullreport.pdf.

13 Mary Jo Pugh, “The Illusion of 18 William J. Maher, “Use of User Studies,” Omniscience: Subject Access and the Midwestern Archivist 11 (1986): 21. Reference Archivist,” American Archivist 19 Jane Stevenson, “’What Happens If 45, no. 1 (winter 1982): 38. Dowler also I Click on This?’ Experiences of the targeted mediation with humor: “To Archives Hub,” Ariadne 57 (October archivists, mediation has generally 2008). Published online at: http://www. meant the satisfying vision of the erudite ariadne.ac.uk/issue57/stevenson/. archivist leading a grateful scholar by the hand through the uncharted forest of 20 Northwest Digital Archives (NWDA), records to precisely the right material.” “Executive Summary: Usability Test #5” “The Role of Use in Defining Archival (November 6, 2008), 6. Published online Practice and Principles,” American at: http://www.orbiscascade.org/index/ Archivist 51 (winter/spring 1988): 82. usability-design-working-group-reports.

14 For example, Jennifer Ward, Pam Mojfeld, 21 NWDA, “Executive Summary: Usability and Steve Shadle, “WorldCat Local at Testing Round 4” (March 12, 2008), 6, 7. the University of Washington Libraries,” 22 Wendy Duff, and Penka Stoyanova, Library Technology Reports 44, no. 6 “Transforming the Crazy Quilt: Archival (August /September 2008). Published Displays from the Users Point of View,” online at: http://www.techsource.ala. Archivaria 45 (spring 1998): 63. “Many of org/ltr/-local-at-the-univ- the historians interviewed, particularly of-washington-libraries.html. the social historians, suggested that 15 Ross Housewright and Roger Schoenfeld, subject indexes, keyword searches, or Ithaka’s 2006 Studies of Key Stakeholders identification of themes would help in the Digital Transformation in Higher them with their research”: Wendy Duff Education (August 2008). Published online and Catherine Johnson, “Accidentally at: http://www.ithaka.org/research/ Found on Purpose: Information-Seeking Ithakas%202006%20Studies%20 Behavior of Historians in Archives,” of%20Key%20Stakeholders%20in%20 Library Quarterly 72, no. 4 (2002): 490. the%20Digital%20Transformation%20 23 21% were by name, and 33% were by in%20Higher%20Education.pdf. the type of document. Louise Gagnon- 16 Sarah Shatforth (Layne) first applied Arguin, “Les questions de recherché the concepts Ofness and Aboutness comme matériau d’études des usagers to cataloging: “Analyzing the Subject en vue du traitement des archives,” of a Picture: a Theoretical Approach,” Archivaria 46 (winter 1998): 92. Cataloging & Classification Quarterly 24 Mary Jo Pugh, “The Illusion of 6 (1986): 39-62. Users and user Omniscience: Subject Access and testers have not used the terms. the Reference Archivist,” American Archivist 45, no. 1 (winter 1982): 40.

93 Making Archival and Special Collections More Accessible

25 Dana Chisnell, Report: RLG Archival 33 Here more than one answer was Resources Service Redesign Focus Groups possible, hence a total of over 100%. (Usability Works, 2 December 2004), 18. Susan Hamburger, “How Researchers For similar results from recent user studies Search for Manuscript and Archival in a library context, see Karen Calhoun Collections,” Journal of Archival et al., Online Catalogs: What Users and Organization 2, no. 1/2 (2004): 84. Librarians Want, (Dublin, Ohio: OCLC, 34 Christopher Prom, “User Interactions 2009). Published online at: http://www. with Electronic Finding Aids in a oclc.org/reports/onlinecatalogs/fullreport. Controlled Setting,” American Archivist pdf. See also 2004 Information Format 67, no. 2, (fall/winter 2004): 263. Trends: Content, Not Containers (Dublin, Ohio: OCLC, 2004). Published online 35 Helen Tibbo and Lokman I. Meho, at: http://www5.oclc.org/downloads/ “Finding Finding Aids on the World community/2004infotrends_content.pdf. Wide Web,” American Archivist 64, no. 1 (spring/summer): 61. 26 Jackie M. Dooley, “Subject Indexing in Context,” American Archivist 55 (spring 36 California Digital Library, OAC Usability 1992): 351. Richard Smiraglia proposed Test Summary (October 2001), 1. methodology for subject access in 37 Northwest Digital Archives, “Executive “Subject Access to Archival Materials Summary: Usability Test #5” (November Using LCSH,” Cataloging & Classification 6, 2008), 2. Published online at: http:// Quarterly 11, no. 3/4 (1990): 63-90. www.orbiscascade.org/index/usability- 27 Arnold Arcolio and Felicia Poe, design-working-group-reports. WorldCat Local at the University 38 Kristina Southwell, “How Researchers of California: Usability Testing, Learn of Manuscript Resources Round One (Spring 2008), 4. in Western History Collections,” 28 Michelle Light, “The Endangerment Archival Issues 26, no. 2 (2002):99. of Trees,” EAD @ 10 (August 31, 2008): 39 Susan Hamburger, “How Researchers 1, forthcoming. Published online at: Search for Manuscript and Archival http://www.archivists.org/publications/ Collections,” Journal of Archival proceedings/EAD@10/[email protected]. Organization 2, no. 1/2 (2004): 91. 29 Lawrence Dowler quotes Richard Szary 40 Carole L. Palmer, Lauren C. Teffeau saying that provenance information and Carrie M. Pirmann, Scholarly and content indexing are each half of an Information Practices in the Online approach to improving description and Environment: Themes from the Literature retrieval: “The Role of Use in Defining and Implications for Library Service Archival Practice and Principles,” American Development (Dublin, Ohio: OCLC Archivist 51 (winter/spring 1988): 83. Research, 2009): 10. Published online 30 Wendy M. Duff and C.A. Johnson, at: http://www.oclc.org/programs/ “Accidentally Found on Purpose: publications/reports/2009-02.pdf. Information-Seeking Behavior of 41 Jihyun Kim, “EAD Encoding and Display: Historians in Archives,” Library A Content Analysis,” Journal of Archival Quarterly 72, no. 4 (2002): 472-496. Organization 2, no. 3 (2004): 50. 31 NWDA, “Executive Summary: Usability 42 Wendy Duff and Catherine Johnson, Testing Round 4” (March 12, 2008), 6, 7. “Accidentally Found on Purpose: 32 E-mail correspondence 12 November 2008. Information-Seeking Behavior of Historians in Archives,” Library Quarterly 72, no. 4 (2002): 493-494.

94 The Metadata is the Interface: Better Description for Better Discovery

43 Neither users nor testers ever suggest 52 Joanne Archer, Ann M. Hanlon and that Library of Congress Subject Jennie Levine, “Investigating Primary Headings (LCSH) are what they mean Source Literacy,” forthcoming [in by “subjects;” LCSH is never even Journal of Academic Librarianship 35, mentioned by testers or users. no. 5 (September 2009)]: 13. Archer, Hanlon and Levine report students’ 44 Jane Stevenson, “’What Happens If searches, using keywords, returned I Click on This?’: Experiences of the thousands of results and very few Archives Hub,” Ariadne 57 (October students refined or limited their results. 2008). Published online at: http://www. ariadne.ac.uk/issue57/stevenson/. 53 Christopher Prom, “User Interactions with Electronic Finding Aids in a 45 Wendy Duff and Catherine Johnson, Controlled Setting,” American Archivist “Accidentally Found on Purpose: 67, no. 2, (fall/winter 2004): 254. Information-Seeking Behavior of Historians in Archives,” Library 54 Andrea Rosenbusch, “Are Our Users Quarterly 72, no. 4 (2002): 485. Being Served? A Report on Online Archival Databases,” Archives and 46 Sara Snyder et al., [Archives of American Manuscripts 29, no. 1 (2001): 50. Art] Web Usability Study Report: Round 1, Subject Experts, (April 2008). 55 Sara Snyder et al., [Archives of American Art] Web Usability Study Report: Round 47 Dana Chisnell, Report: RLG Archival 1, Subject Experts (April 2008). Resources Service Redesign Focus Groups (Usability Works, 2 December 2004), 14-15. 56 Jane Stevenson, “’What Happens If I Click on This?’: Experiences of the 48 Richard Lytle, “Intellectual Access to Archives Hub,” Ariadne 57 (October Archives: I. Provenance and Content 2008). Published online at: http://www. Indexing Methods of Subject Retrieval,” ariadne.ac.uk/issue57/stevenson/. American Archivist 43 (winter 1980): 68. 57 Dana Chisnell, ArchiveGrid Rapid 49 Elsie T. Freeman, “In the Eye of the Iterative Testing Report (Usability Beholder: Archives Administration from Works, 28 February 2006), 12, 17. the User’s Point of View,” American Archivist 47, no. 2 (spring 1984): 116. 58 Wendy Duff, and Penka Stoyanova, Wendy Duff and Catherine Johnson “Transforming the Crazy Quilt: Archival added the requirement that systems Displays from the Users Point of View,” be intuitive: “Accidentally Found on Archivaria 45 (spring 1998): 57. Purpose: Information-Seeking Behavior 59 Andrea Rosenbusch, “Are Our Users of Historians in Archives,” Library Being Served? A Report on Online Quarterly 72, no. 4 (2002): 472. Archival Databases,” Archives and 50 “[R]ecords in an archives are not Manuscripts 29, no. 1 (2001): 49. records about an activity, but records 60 Michelle Light, “The Endangerment of an activity.”[Emphasis added.] For of Trees,” EAD @ 10 (August 31, 2008): a concise summary of how historians 2; forthcoming. Published online at: use provenance information for subject http://www.archivists.org/publications/ queries, see Wendy Duff and Catherine proceedings/EAD@10/[email protected]. A. Johnson, “Accidentally Found on Purpose: Information-Seeking Behavior of 61 NWDA, “Executive Summary: Usability Historians in Archives,” Library Quarterly Testing Round 4” (March 12, 2008), 4. 72, no. 4 (2002): 477, including footnote 5.

51 Avra Michelson, “Description and Reference in the Age of Automation,” American Archivist 50 (spring 1987): 199.

95 Making Archival and Special Collections More Accessible

62 et al., “Creating the Next 70 Erika Dowell, “Web Site Usability for Generation of Archival Finding Aids,” D-Lib Rare Book and Manuscript Libraries,” 13, no. 5/6 (May/June 2007). Published RBM 9, no. 2 (fall 2008): 175, 179. online at: http://www.dlib.org/dlib/may07/ 71 Research Information Network, yakel/05yakel.html. For an intriguing Discovering Physical Objects: Meeting example of possibilities for one specific Researchers’ Needs (London, finding aid, see Magia Ghetu Krause and October 2008): 5. Published online Elizabeth Yakel, “Interaction in Virtual at: http://www.rin.ac.uk/objects. Archives: The Polar Bear Expedition Digital Collections Next Generation 72 Mark A. Greene and Dennis Meissner, Finding Aid,” American Archivist 70, “More Product, Less Process: no. 2 (fall/winter 2007): 282-314. Revamping Traditional Archival Processing,” American Archivist 68, 63 Avra Michelson, “Description no. 2 (fall/winter 2005): 208-263. and Reference in the Age of Automation,” American Archivist 73 Joshua Ranger, “Mass Digitization of 50 (spring 1987): 192, 200. Archival Manuscripts,” (2008). Published online at: http://www.pacsclsurvey. 64 Karen Markey, “Twenty-Five Years of org/documents/ranger/04ranger.ppt. End-User Searching, Part 2: Future Research Directions,” Journal of the 74 Mark Greene, “[Draft Grant Report American Society for Information Science to NHPRC:] Basic Processing User and Technology, 58 no. 8 (2007): 1126. Survey Overview,” (2008).

65 An outstanding counter-example is 75 For example, RLG learned this when Jody DeRidder’s use of sitemaps and MARC records were mixed with archival links in static html pages to increase finding aids. See Dana Chisnell, dramatically discovery of finding aids at Report: RLG Archival Resources Service the University of : “Googlizing Redesign Focus Groups, (Usability a Digital Library,” The Code4Lib Journal Works, 2 December 2004), 18. 2 (March 24, 2008). Published online at: 76 Joanne Archer, Ann M. Hanlon and http://journal.code4lib.org/articles/43/. Jennie Levine, “Investigating Primary 66 Cathy De Rosa et al., College Students’ Source Literacy,” forthcoming [in Perceptions of Libraries and Information Journal of Academic Librarianship Resources (Dublin, Ohio: OCLC, 2006), 1-1. 35, no. 5 (September 2009)]: 17. Published online at: http://www.oclc.org/ 77 Wendy Duff, and Penka Stoyanova, reports/pdfs/studentperceptions.pdf. “Transforming the Crazy Quilt: Archival 67 Joanne Archer, Ann M. Hanlon and Displays from the Users Point of View,” Jennie Levine, “Investigating Primary Archivaria 45 (spring 1998): 52-54, 59. Source Literacy,” forthcoming [in 78 Dana Chisnell, ArchiveGrid Rapid Journal of Academic Librarianship Iterative Testing Report (Usability 35, no. 5 (September 2009)]: 11. Works, 28 February 2006), 11. 68 Christopher Prom, “User Interactions 79 Jane Lee, OAC First round Usability with Electronic Finding Aids in a Test Findings (OAC Redesign Project, Controlled Setting,” American Archivist 11 September 2008), 15. 67, no. 2, (fall/winter 2004): 247. 80 NWDA, “Executive Summary: Usability 69 Jane Stevenson, “’What Happens If Testing Round 4” (March 12, 2008), 6. I Click on This?’: Experiences of the Archives Hub,” Ariadne 57 (October 81 Dana Chisnell, Report: RLG Archival 2008). Published online at: http://www. Resources Service Redesign Focus Groups ariadne.ac.uk/issue57/stevenson/. (Usability Works, 2 December 2004), 17.

96 The Metadata is the Interface: Better Description for Better Discovery

82 RLG in-house Archival Resources 90 Maher, William J., “Use of User Studies,” correspondence from 2003. See also Midwestern Archivist 11 (1986): 22. Cathy De Rosa et. al., Perceptions 91 Jane Stevenson, “’What Happens If of Library and Information Resource I Click on This?’: Experiences of the (Dublin, Ohio: OCLC Research, 2005), ix. Archives Hub,” Ariadne 57 (October Published online at: http://www.oclc. 2008). Published online at: http://www. org/reports/pdfs/Percept_all.pdf. ariadne.ac.uk/issue57/stevenson/. 83 “Generally, participants did not talk about 92 Kathleen Feeney, “Retrieval of tools for doing research…It is unlikely Archival Finding Aids Using World- that researchers approach doing research Wide-Web Search Engines,” American by looking for a tool for doing research.” Archivist 62:2 (1999): 224. Dana Chisnell, Report: RLG Archival Resources Service Redesign Focus Groups 93 Spindler, Robert P. and Richard Pearce- (Usability Works, 2 December 2004), 13. Moses, “Does AMC Mean ‘Archives Made Confusing? Patron Understanding 84 Karen Markey, “Twenty-Five Years of of USMARC AMC Catalog Records,” End-User Searching, Part 1: Research American Archivist 56 (spring 1993): 340. Findings,” Journal of the American Society for Information Science and 94 Michelle Light, “The Endangerment Technology, 58 no. 8 (2007): 1071. of Trees,” EAD @ 10 (August 31, 2008): 3-5, forthcoming. Published online at: 85 “Le traitement des fonds demeure http://www.archivists.org/publications/ cependant la préoccupation des centres proceedings/EAD@10/Light-EAD@10. d’archives et le lien entre l’information pdf. Dryden is also concerned that we décrite et la clé d’accès fournie est have—for years—developed standards for d’autant plus fragile dans un contexte archival description without consulting informatique,” Louise Gagnon-Arguin, “consumers” of these descriptions: Jean “Les questions de recherché comme Dryden, “Do We Care What Users Want? matériau d’études des usagers en Evaluating User Satisfaction and the vue du traitement des archives,” LibQUAL+™ Experience,” Journal of Archival Archivaria 46 (winter 1998): 100. Organization 4, no. 4 (2004): 84-85. 86 Dana Chisnell, Report: RLG Archival 95 Lorcan Dempsey says that his aphorism, Resources Service Redesign Focus Groups “Discovery happens elsewhere,” “… (Usability Works, 2 December 2004), 13. is influenced by Steve Rubel’s use of 87 Research Information Network, the phrase, ‘traffic happens elsewhere’ Discovering Physical Objects: Meeting in discussion of the ‘Cut and Paste Researchers’ Needs (London, Web’ < http://www.micropersuasion. October 2008), 5. Published online com/2007/08/the-cut-and-pas. at: http://www.rin.ac.uk/objects. html >. Published online at: http:// orweblog.oclc.org/archives/001430. 88 Elizabeth Yakel, “Listening to Users,” html (September 16, 2007). Archival Issues 26:2 (2002): 118. Susan Hamburger, “How Researchers 96 William J. Maher, “Use of User Studies,” Search for Manuscript and Archival Midwestern Archivist 11 (1986): 17. Collections,” Journal of Archival Organization 2, no. 1/2 (2004): 83, 89. William J. Maher, “Use of User Studies,” Midwestern Archivist 11 (1986): 23.

89 Kristina Southwell, “How Researchers Learn of Manuscript Resources in Western History Collections,” Archival Issues 26, no. 2 (2002):100.

97 Making Archival and Special Collections More Accessible

98 Making Archival and Special Collections More Accessible 8 “Capture and Release”: Digital Cameras in the Reading Room

Lisa Miller, Steven K. Galbraith, and the RLG Partnership Working Group on Streamlining Photography and Scanning

This paper was originally published in February 2010 by OCLC Research at http://www.oclc.org/content/dam/research/publications/library/2010/2010-05.pdf.

99 Making Archival and Special Collections More Accessible

Introduction

Digital cameras are revolutionizing special of a broader universe of materials, like oversize collections reading rooms and the research materials and bound volumes that are excluded process, much as photocopy machines did from the photocopy policies of many repositories, for a previous generation. Reference routines and they can make color copies. Given a choice focused on the photocopier are embedded in between two repositories, one that has more workflows of every repository; photocopying generous policies and one that does not, is accepted by repositories, tolerated by rights researchers may make choices accordingly. holders, and expected by researchers. Now Digital cameras reduce repository workload— technology is forcing repositories to confront Depending on the repository’s photocopy and change again. The ubiquity of digital cameras digital camera policies, allowing personal digital and other mobile capture devices has resulted cameras outsources duplication tasks to the in researchers desiring and expecting to use user, freeing staff to perform other work in these cameras in reading rooms. While some librarians times of increased demands, expectations, and and archivists have resisted digital cameras, workloads. In addition, cameras may reduce others have embraced them—and rightfully photocopier maintenance and supplies. so. The benefits to researchers, repositories, and collection materials are undeniable. Digital cameras enhance security and save reading room checkout time—Digital cameras decrease the number of photocopies leaving Benefits the reading room in the hands of researchers, reducing checkout time and the opportunity Digital cameras are gentler on collection for theft. With twentieth- and twenty-first- materials—Upending collection materials to century collections, it is frequently difficult to position them on a photocopy machine, even distinguish between copies and originals. when done with the utmost care, risks more damage to materials than photographing Digital cameras save paper and photocopy them in the reading room while they are toner—Photographing materials is an effortless face up and appropriately supported. The way to reduce our environmental impact. materials are not subjected to the intense Repositories stay current and resolve an light of a photocopier, but rather are usually ongoing issue—Repositories remain largely easily photographed with ambient lighting. analog outposts, in contrast to the 24/7 online Digital cameras facilitate use—Researchers with world that most people live and work in. As much limited time can cover more collection materials as we would like to deliver collection materials during their visit by photographing relevant to all online, it is still beyond our grasp. Digital materials for in-depth study later. We preserve cameras are research tools that reach across this these materials so that they can be used. More use online/offline divide, one researcher at a time. allows us to report higher reference figures and Digital cameras reduce liability for significant research use to our resource allocators. infringement—Digital cameras lessen the Digital cameras increase researcher repository’s risk profile, especially if it maintains satisfaction—Researchers must take time a “hands-off” approach towards the use of from work and school to travel to our reading personal cameras. When a repository makes rooms during our limited business hours, often copies of copyrighted documents for users or at great expense. Just as libraries and archives provides equipment on which users can make struggle with tighter budgets in these challenging their own copies, it runs the risk of engaging in economic times, so, too, do researchers. Digital direct and indirect copyright infringement. cameras maximize their precious time in the reading room and end their wait for copies. Depending on the nature of the repository’s Duplication, Copyright camera use policy, patrons may also save money and eliminate time spent on photocopy and the Web request paperwork. They may also make copies 100 "Capture and Release": Digital Cameras in the Reading Room

Section 108 of U.S. copyright law allows Suggested Practices for repositories to make digital copies of textual material for private study, scholarship, or Cameras in the Reading Room research. If making a copy of an entire book or To synthesize a core of suggested practices, manuscript item or a substantial part of it for the RLG Partnership working group reviewed a user, the repository must determine that a the current policies of thirty-five repositories copy is not available at a fair price.1 For non- comprised of academic libraries, independent textual material, such as photographs, the law research libraries, historical societies, allows repositories to make copies only if the government archives, and public libraries repository concludes that the user’s request (see Appendix A. Policies Reviewed). Below is a fair use—a difficult, and potentially risky, are the most commonly shared elements, assumption.2 By allowing patrons to use their arranged in categories for administration digital cameras, the repository removes itself and handling of collection materials. from the duplication process and eliminates the risk associated with making copies. Section 108(f)(1) protects a repository from Administration secondary liability for the “unsupervised use of • Require camera users to complete and reproducing equipment located on its premises” sign an application/policy/terms-of-use (emphasis added), provided that the equipment form agreeing that images of sensitive and displays a notice that making copies may be copyrighted materials will only be used for 3 subject to copyright law. Ironically, supervised study, teaching, or research purposes and use of reproduction equipment, such as requiring will be used in compliance with copyright users to seek permission before making any law. Some agreements also stipulate that copies, increases the repository’s risk of liability. the user cannot reproduce images without Rather than place a notice on cameras, the permission from the institution. A few forms Section 108 Study Group recommended that require the user to list specifically what he a notice be posted prominently in public or she is digitally reproducing. This allows areas stating that making copies may be the institution to keep statistics on what and subject to copyright law.4 Such a statement how much is being digitized and to check should also appear on digital camera use whether any of the materials already exist in agreements signed by researchers. the institution’s digital repository, though it increases liability for copyright infringement. Reading room photography does not lead inexorably to collection materials inappropriately • Staff reviews collection materials prior ending up online. This issue is already to photography. This ensures that items managed by each repository’s publication are not too fragile to be reproduced and policy. Many repositories have been providing allows staff to note any copyright or donor digital reproductions to patrons for years restrictions, though it also places the under existing duplication and publication institution at greater risk of liability. policies. Given how easy it is to digitize analog • Limit the number of shots, when reproductions, drawing distinctions between appropriate, to a quantity determined analog and digital copies makes little sense. by institutional policy and/or in Some repositories consider responsible reuse accordance with copyright policies. of images on the Web as good outreach. • Watermark digital reproductions by requiring that each item be photographed with a streamer, transparency, or card that identifies the item and its holding institution and, if applicable, displays a copyright notice. Patrons are responsible for properly citing their copies, but repositories may provide citation guidelines. • Digital photography must not disturb other users or staff. All audio functions 101 Making Archival and Special Collections More Accessible

on digital cameras must be turned off Evolving Practices for Digital and users may not photograph other patrons, staff, or the reading room. Cameras in Reading Rooms • Beyond the suggested practices above, Handling Collection Materials many facets of digital camera use continue to develop and can be implemented • No flash photography. It is a independently along sliding scales distraction to other users. represented in Table 1. A repository can mix • As with any method of duplication, camera and match from these modules according use is considered only if it will not damage to its nature, needs, and inclination. collection materials. Users are instructed on • Established photocopy policies and how to handle items during photography. processes often form the baseline for • In an effort to monitor how users handle a repository’s digital camera policy. If items during photography, several the staff performs all photocopying, an institutions designate specific work areas appointment and designated workstation where items may be photographed or have for digital photography, supervised by the a staff member present during shooting. photocopy staff and with time charged to Some provide or require use of an in- the researcher, may be the logical approach. house camera stand. Some policies make As an alternative, the digital camera policy a point of prohibiting users from bringing could steer researchers toward some goal their own tripods or lighting equipment. of the repository, such as reducing the staff’s photocopy workload or achieving a paperless duplication system. A repository might encourage both of these goals by continuing to charge its standard rate for photocopies while not charging for copies made with digital cameras. The repository can swap out one facet for another as it experiments with cameras, and gradually settle on a policy that works for it.

102 "Capture and Release": Digital Cameras in the Reading Room

Table 1. Faceted Camera Use Grid

Facet Shutter-bug Exposed Camera-shy Traditional Self-service Self-service after All copying done photocopying (possible staff review by staff baseline for digital camera policy)

Equipment • No flash, no lights • Patron’s camera Repository’s camera • Allow flatbed • Limits on supporting only (and possibly scanners equipment (copy other equipment supplied by repository) • Allow and/or provide stands, tripods, copy stand, tripod, cords, etc.)—some extension cords, pieces allowed, stepstool, etc. others not • Repository supplies • No flatbed scanners camera or self- service overhead book scanner in addition to allowing patron’s camera

Photography space In reading room In reading room at Separate room at any station designated stations, usually close to

Photography rules • No standing on tables or chairs • No rearrangement of furniture • No materials on floor • Remain behind table, facing forward at all times • Set camera to “mute” • Do not disturb others • No photographs of reading room, staff, or patrons

Appointments Appointment Appointment made Appointment made in not required during visit writing in advance

Staff review of Part of standard staff Patron must verbally • Patron must collection materials surveillance of patrons notify reference formally indicate in reading room attendant each time and curatorial staff camera is used and formally review show attendant all materials the materials being • Camera stays photographed in locker until approval is given • Same-day approval may not be possible

103 Making Archival and Special Collections More Accessible

Facet Shutter-bug Exposed Camera-shy Materials • Preservation needs always trump photography needs handling rules • Do not manipulate materials to achieve a better image • Do not press down on materials or bindings • Manuscript materials must always be flat on the table and not held up in air • Loose materials must remain in their folder and in order at all times • Photograph materials from one folder at a time • Volumes should not be laid flat—book cradles will be provided • Weight bags and snakes are available • Do not fold pages • Do not remove fasteners—ask for staff assistance • Do not remove items from sleeves, mats, etc.

Quantity limits No limits • No more than • Limit to established 50 pages or 20 number of shots percent (whichever per day is smaller) of any • Patron’s images may manuscript or book be reviewed during • No entire book, checkout to enforce manuscript box, quantity limit or collection, nor substantial portions of them • Please limit number of photographs to a reasonable amount • Photographs are meant to alleviate photocopying and supplement note taking, not to create a complete personal copy

Other limits • Oversize items or anything that does not safely fit on table • Fragile or damaged items • No materials received on interlibrary loan, unless permits • Only materials checked out to the patron using the camera • If not allowed, staff may digitize at standard fees • Repository reserves right to deny permission for any collection materials at its discretion

104 "Capture and Release": Digital Cameras in the Reading Room

Facet Shutter-bug Exposed Camera-shy Copyright • Copyright notice (and citation) in all shots (paper strip or transparency) • Digital copies are for personal research use only • Repository displays a copyright warning where digital camera requests are accepted and on digital camera policy forms

Paperwork (in addition • Camera use Patron provides • Patron provides to forms completed agreement included list of collections list of each item by all patrons) on registration form (plus camera use (plus camera use • Separate camera agreement) agreement) use agreement with • Written request copyright declaration before visit (renewed annually/ per visit/per day)

Fees None • Minimal fee (per Fee equals or exceeds visit, per shot) cost of photocopies • Fees for equipment supplied by repository

Publication Publication requires Images for publication • Images taken by permission of the or distribution must patron may not be copyright holder be ordered through published in print the library at set fees or on Internet • Publication requires written permission from repository

Citations • Patron is responsible for recording complete citations for each shot • Subsequent orders for high-resolution images cannot be processed without complete citations • Source repository template in all shots (paper strip or transparency, often included with copyright notice)

Other • Camera privileges can be revoked at any time if rules are not followed • Provide tips on taking good images and creating complete citations • In some cases, repository receives copies of all photographs

105 Making Archival and Special Collections More Accessible

Many repositories charge for photocopies and Conclusion this can form a basis for charging for digital camera use. When a repository establishes such Digital cameras are the newest research tool, but fees, it should follow the guidelines noted in the they will not be the last. The next generation of ALA/SAA Joint Statement on Access to Research archivists, librarians, and curators will view digital Materials in Archives and Special Collections cameras the way we currently view photocopy Libraries. This document states in part, “A machines, as essential components of our repository should facilitate access to collections reference system. The issues of new technology by providing reasonably priced reproduction are wrongly framed as a threat or a challenge services that are administered consistently for repositories to remain relevant.8 Rather, in accordance with legal authority, including digital cameras should be considered from the copyright law, institutional access policy, and perspective of our most fundamental goals— repository regulations. These services . . . should improving conditions for our collections materials, be clearly stated in a publicly accessible written facilitating greater research economically and policy.”6 Charging fees for reproductions of efficiently, and resolving competing demands copyrighted material may place the institution for resources and maximizing the productivity at greater risk for copyright infringement. If the of our staff. By adopting this mindset with our fees are determined to provide “direct or indirect digital camera policies, we are poised to evaluate commercial advantage” to the repository, its objectively the technology that will replace digital Section 108 exemptions are lost and maintaining cameras in the next generation—or sooner. a “fair use” defense becomes much harder. Peter Hirtle, Jim Kuhn, Merrilee Proffitt, Jackie A few repositories have introduced particularly Dooley and Ricky Erway reviewed early versions of unique facets to their digital camera policies, this report. The final document is better as a result as noted in the “other” section of the grid. of their comments, which are greatly appreciated. Some ask for copies of all digital images, with citations, and add them to the repository’s collection of digital assets. In these cases, the repository may wish to include a statement to that effect in the digital camera use agreement. To assist researchers in obtaining usable photographs and citations, some repositories provide photography tips to their patrons.7

106 "Capture and Release": Digital Cameras in the Reading Room

Appendix A. Policies Reviewed

• American Antiquarian Society • University, Fales Library • Arizona State University, Arizona • The Newberry Library Historical Foundation • San Francisco Public Library, San • , L. Tom Francisco History Center Perry Special Collections • Stanford University, Hoover • California Historical Society Institution Archives • , Division of Rare • Stanford University, Special Collections and Manuscript Collections and University Archives • Dallas Theological Seminary • Syracuse University • Duke University, Rare Book, Manuscripts, • Tulane University, and Special Collections Library Research Collection • Emory University, Pitts Theology Library • University of Alaska Anchorage & Alaska Archives and Manuscripts Department Pacific University Consortium Library, Archives & Special Collections • Folger Shakespeare Library • University of California, Berkeley, • Frick Art Reference Library Robbins Collection • Getty Research Institute • University of California, Irvine, Langson • Harvard University, Houghton Library Library Special Collections • The Huntington Library, Arts • University of California, Los Angeles, Collections, and Botanical Gardens Charles E. Young Research Library • Indiana University Bloomington, Lilly Library Department of Special Collections • Library of Congress, Prints & • University of Maryland at College Photographs Division Park, Special Collections • Library of Virginia • University of Miami, Special Collections and University Archives • Minnesota Historical Society • University of Texas at Austin, • The National Archives at The Harry Ransom Center College Park, Maryland • University of Virginia, Albert and Shirley • The National Archives, United Kingdom Small Special Collections Library • New York Public Library, Manuscripts and Archives Division, Berg Collection, and Schomburg Center

107 Making Archival and Special Collections More Accessible

Appendix B. Draft Modular Form: Camera Use in Reading Room

This sample form can be adapted by a repository by deleting irrelevant sections or inserting additional specific requirements. It is available as a standalone editable document on the OCLC Research Web site at http://www.oclc.org/research/activities/photoscan/policy.doc.

Camera Use Policy Researchers may take photographs of collection materials for study purposes only, and as allowed by the library, based on the physical condition of the materials, copyright law, donor restrictions, and reading room rules. I agree to the following conditions: Repository procedures [delete or add as needed] • I will obtain permission from library staff before taking any photographs. • I will indicate all items to be photographed and show them to library staff for approval. • I will provide a list of all [collections or items] photographed. • I will take photographs at designated stations only. • I will not photograph more than [50 pages or 20 percent of any book or manuscript (whichever is smaller), 100 pages per collection, other arbitrary limit]. • I will use my personal camera only—not portable scanners, [phone cameras, other]. • I will include in each photograph a strip provided by the library stating [repository name, copyright notice, and/or citation]. • It is my responsibility to keep accurate citations for all items photographed, which I will need when ordering publication-quality images or requesting permission to quote. Materials handling rules [delete or add as needed] • I will handle the materials with care and according to library rules. • I will not bend, press down, or otherwise manipulate or rearrange materials to get a better photograph. • I will keep materials flat on the table or in the stand/cradle provided. • I will ask library staff for assistance with fastened items. • I will not remove items from their plastic sleeves. • I will not stand on chairs, tables, or other furniture. • I will turn off the flash and sound on my camera. • I will not use special lights [other prohibited equipment]. • I will not take photographs of the staff, reading room, or other researchers. • I understand that the library reserves the right to deny permission to photograph collection materials at its discretion.

108 "Capture and Release": Digital Cameras in the Reading Room

Copyright [delete or add as needed] • I will use the photographs for my private study, scholarship and research only. • I will not publish the photographs in print, post them on the Internet, nor exhibit them. • I will not donate, sell, or provide the photographs to another repository. • I will request publication-quality images from the library at its standard fees. • It is my responsibility to obtain permission to publish from copyright owners. Repository gets copies of all photographs taken [delete or add as needed] • I will provide copies of all of my photographs and citations to the library, and I assign any intellectual property rights that I may possess in them to the repository. WARNING CONCERNING COPYRIGHT RESTRICTIONS The copyright law of the United States (Title 17, United States Code) governs the making of photocopies or other reproductions of copyrighted material. Under certain conditions specified in the law, libraries and archives are authorized to furnish a photocopy or other reproduction. One of these specified conditions is that the photocopy or reproduction is not to be “used for any purpose other than private study, scholarship, or research.” If a user makes a request for, or later uses, a photocopy or reproduction for purposes in excess of “fair use,” that user may be liable for copyright infringement. This institution reserves the right to refuse a copying order if, in its judgment, fulfillment of the order would involve violation of copyright law. I agree to indemnify and hold harmless [repository name], its agents and employees against all claims, demands, costs and expenses incurred by copyright infringement or any other legal or regulatory cause of action arising from the use of these photographs. I have read and agree to abide by the terms and conditions above. I understand that my failure to follow them may result in the termination of my camera privileges. ______Signature Date ______Name (Please print)

List of collections photographed. Please print clearly. [delete or add as needed]

1.

2.

List of items photographed. Please print clearly. [delete or add as needed]

Item Collection Box Folder Item Description

1

2

109 Making Archival and Special Collections More Accessible

Appendix C. Members of the RLG Partnership Working Group on Streamlining Photography and Scanning

• Anne Blecksmith • Elizabeth McAllister Getty Research Institute University of Maryland • Eleanor Brown • Lisa Miller Cornell University Hoover Institution Library and Archives, Stanford University • Paul Constantine University of Washington • Timothy Pyatt Duke University • Gordon Daines Brigham Young University • Jennifer Schaffner OCLC Research • Tiah Edmunson-Morton Oregon State University • Shannon Supple Robbins Collections • Cristina Favretto University of California, Berkeley University of Miami • Francine Snyder • Steven K. Galbraith Solomon R. Guggenheim Museum Folger Shakespeare Library • Mattie Taormina • Susan Hamson Stanford University Columbia University • Cherry Williams • Sue Kunda Lilly Library, Indiana University Bloomington Oregon State University • Jennie Levine Knies University of Maryland • Suzannah Massen Frick Art Reference Library • Dennis Massie OCLC Research • Dennis Meissner Minnesota Historical Society

110 "Capture and Release": Digital Cameras in the Reading Room

Notes

1 “Limitations on exclusive rights: 7 Lyndon Baines Johnson Library and Reproduction by libraries and archives,” Museum, “Instructions for Use of Digital Section 108, Chapter 1, Title 17, United Cameras in the Reading Room,” May 20, States Code, accessed February 1, 2010, 2004, accessed February 1, 2010, at http:// at http://www.law.cornell.edu/uscode/17/ www.lbjlib.utexas.edu/johnson/archives. usc_sec_17_00000108----000-.html. hom/digital_camera_info.shtm, and Kirklin Bateman, Sheila Brennan, Douglas 2 For more on direct and indirect copyright Mudd, and Paula Petrik, “Taking a Byte infringement and its relevance for reading Out of the Archives: Making Technology room practice, see: Peter B. Hirtle, Work for You,” Perspectives, 43,1 (January Emily Hudson, and Andrew T. Kenyon, 2005), accessed February 1, 2010, at Copyright and Cultural Institutions: http://www.historians.org/Perspectives/ Guidelines for Digitization for U.S. Issues/2005/0501/0501arc1.cfm. Libraries, Archives, and Museums (Ithaca, NY: Cornell University Library, 2009): 8 Richard Cox with the University of pp 78-83, accessed February 1, 2010, Pittsburgh archives students, “Machines in at http://hdl.handle.net/1813/14142. the Archives: Technology and the Coming Transformation of Archival Reference,” 3 “Limitations on exclusive rights: First Monday, 12,11 (November 2007), Reproduction by libraries and archives,” accessed February 1, 2010, at http://www. Section 108, Chapter 1, Title 17, United uic.edu/htbin/cgiwrap/bin/ojs/index. States Code, accessed February 1, 2010, php/fm/article/viewArticle/2029/1894. at http://www.law.cornell.edu/uscode/17/ usc_sec_17_00000108----000-.html.

4 Copyright Office and National Digital Information Infrastructure and Preservation Program of the Library of Congress, Section 108 Study Group Report 91-92 (2008), accessed February 1, 2010, at http://section108.gov/docs/ Sec108StudyGroupReport.pdf.

5 For a study of the effects of light on various materials, see: Terry T. Schaeffer’s Effects of Light on Materials in Collections: Data on Photoflash and Related Sources (Los Angeles, CA: Getty Publications 2001).

6 ALA/SAA Joint Statement on Access to Research Materials in Archives and Special Collections Libraries, adopted by the SAA Council on June 1, 2009, and the ACRL Board during the ALA Annual Conference, July 2009, accessed February 1, 2010, at http://www.archivists.org/ statements/ALA-SAA-Access09.asp and http://www.ala.org/ala/mgrps/divs/ acrl/standards/jointstatement.cfm.

111 Making Archival and Special Collections More Accessible

112 Making Archival and Special Collections More Accessible 9 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Dennis Massie Program Officer OCLC Research

This paper was originally published in July 2013 by OCLC Research at http://www.oclc.org/content/dam/research/publications/library/2013/2013-03.pdf.

113 Making Archival and Special Collections More Accessible

Acknowledgments

Jennifer Schaffner, Program Officer, OCLC Our working and advisory groups enjoyed a Research, has from the beginning been the special collaborative relationship with the RBMS driving force behind this work. Her passion Task Force responsible for the 2012 revision of for bringing special collections materials ACRL’s Guidelines for Interlibrary and Exhibition to a wider audience continues to be both Loans of Special Collections Materials, thanks to transformative and inspirational. chair Hjordis Halvorson of the Newberry Library, Christian Dupont of Atlas Systems, and our Special thanks go to the members of the Special operative in common, Shannon Supple of UCLA. Collections Delivery Steering Committee— Cristina Favretto of the University of Miami, Members of the 2012 SHARES Executive Group Susan Hamson of Columbia University, and (particularly Lesliediana Jones of the George Mattie Taormina of Stanford University—who in Washington University Law ) 2009 identified the physical sharing of special improved the SHARES Facility Trust Checklist collections as one of a set of key cooperative and added several additional use cases. projects that could transform the way special Ricky Erway, Senior Program Officer, OCLC collections materials are brought together Research, significantly improved both the with the researchers who need them. text and the organization of this report. The Sharing Special Collections Working Group did all the heavy lifting and deserves the lion’s share of the credit for this report: • Jennifer Block, • Scott Britton, Boston College (formerly of University of Miami) • Barbara Coopey, Pennsylvania State University • Aimee Lind, Getty Research Institute • Sandra Stelts, Pennsylvania State University • Dennis Massie and Jennifer Schaffner, OCLC Research The Sharing Special Collections Advisory Group consulted extensively with the working group, guided the conversation about the physical lending of special materials, and served as expert sounding board: • Eleanor Brown, North Carolina State University (formerly of Cornell University) • Laura Carroll, Emory University • Margaret Ellingson, Emory University • Cristina Favretto, University of Miami • Paul Constantine, University of Washington • Suzan Hallgren, University of Minnesota • Elizabeth Nielson, Oregon State University • Shannon Supple, University of California, Los Angeles (formerly of University of California, Berkeley)

114 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Executive Summary This report presents strategies for providing From 2009 through 2011, a working group efficient and affordable interlending of actual made up of resource sharing supervisors physical items from special collections for and special collections curators from OCLC research purposes, as well as advice on Research Library Partnership institutions determining if a loan is the most appropriate studied this issue. The most significant activity way to fulfill a particular request. of the working group was creating a set of tools that will help institutions reconsider and The lending of physical items for exhibition streamline their processes for handling loan purposes has long been a core activity of requests for special collections materials. archivists and special collections curators. Now, with the increased visibility of special collections, These tools include: requests for research loans are multiplying. There • a tiered approach to streamlining are legitimate instances—based on the nature of workflows associated with lending special the material, the type of research question, or the collections, outlining minimal, moderate need for extended access by a distant scholar— and maximum amounts of effort and when only the loan of a physical item from special overhead, to be invoked based on collections can satisfy a researcher’s request. —— the material Prudent approaches to lending rare and unique materials are justified, and providing a digital —— the request surrogate is usually the answer. But such thinking —— the risk tolerance of curators is not appropriate for every item in special and administrators collections, or for every request, and often results in time-consuming, overly cautious procedures. • a model written policy on Streamlining such procedures is critical. Labor- sharing special collections intensive processes and policies can be simplified • a “trust” checklist to serve as a conversation to fit the nature of the material, institutional starter between a prospective lender and resources, the circumstances of requests, and the an institution interested in borrowing risk tolerance of curators and administrators. an item from special collections Lending physical items ranks among the most This report contains a complete description divisive issues in the field of archives and special of the working group’s activities, plus all of collections, perhaps the one most likely to bring the tools listed above, and advice on how out equal parts raw emotion and well-reasoned best to use them. The report’s principles professional opinion. But solid evidence indicates intentionally dovetail with the Association of that the practice of lending physical items from College and Research Libraries’ 2012 revision of special collections is becoming as common as not Guidelines for Interlibrary and Exhibition Loans doing so. While an increasing number of curators of Special Collections Materials (ALA 2012). are willing to consider the physical loan of materials under their stewardship, the workflows for considering and executing such loans tend toward unscalable. In order for curators to cope with the uptick in requests and arrive at a well- considered and professionally-responsible “yes” as often as possible, new workflows and new ways of thinking about lending physical items from special collections must be established.

115 Making Archival and Special Collections More Accessible

Introduction Enhanced discoverability of special collections This report presents strategies for determining if has led to increased interest from researchers. a loan of the original item is the most appropriate Concurrently, advances in scanning technology way to fulfill a particular request for special have helped make the provision of such collections material and offers techniques for materials in digital form fairly routine. There providing efficient and affordable delivery of are instances, however—due to the nature of physical items. Cautious approaches to lending the material, the type of research question, rare and unique materials, while justified, are not or the need for extended access by a distant necessarily appropriate for every item in special scholar—when only the loan of a physical item collections and often result in time-consuming from special collections can satisfy a request. procedures. Labor-intensive processes and policies can be streamlined to fit institutional “Says who?” you might ask. Says two-thirds resources, the circumstances of requests, and the of community practitioners, according to risk tolerance of curators and administrators. a survey conducted for this report. Ten years ago, requests for loans of special items for exhibition purposes were routine. Loans for research purposes were rare. Many institutions refused to consider such requests. Those that did turned each request into what amounted to a special project, requiring multiple internal consultations and extensive contacts between staff at the borrowing and lending institutions. Each step of the process, including packing and unpacking, required the participation of specially-trained experts. With the increased visibility of special collections, requests for physical loans have multiplied. They arrive at prospective lending institutions in two separate streams, directly to the special collections curators and also via interlibrary loan departments. While an increasing number of curators are willing to consider the physical loan of materials under their stewardship, the workflows for considering and executing such loans don’t scale well. In order for curators to cope with the increased volume in requests and arrive at a professionally-responsible “yes” as often as possible, new workflows and new ways of thinking about lending physical items from special collections must be established.

116 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Let’s Get Physical The work described in this report got its initial In 2003, following a well-received Research spark from the same steering committee that Libraries Group program called Sharing the previously championed allowing cameras Wealth, where staff from dozens of institutions in in the reading room and providing scan-on- the US and a few from the UK came together in demand services for users of special collections Washington, D.C., to talk about their experiences materials—both topics, in those days (2009), with sharing physical items from special rather controversial ideas in themselves. But this collections, I formed a working group to develop issue always stood apart. The physical lending a pilot project that would promote such loans. of special collections was put on the table, Within a few weeks, I was pulled aside by the whisked off, and then nudged back on again. “I director of a top-tier ARL library who said, only know we should be talking about this,” said one half-jokingly, “Can’t you find something else to committee member, “if only because it makes work on? This sharing special collections business me feel so uncomfortable.” Another agreed: “We has my staff yelling at each other in the hallways.” have a professional responsibility to push at our The working group’s only UK representatives boundaries and question our comfort zones.” soon begged off, because they felt their peers The third added, “I love the idea of lending were not ready for a rational conversation about from special collections. But I would never be the topic. The working group ended up gathering allowed to bring it up at my own institution.” some interesting examples of documentation and best practices for lending special collections, The idea has been around for a long time. Some but the pilot project itself never materialized. prestigious institutions have been doing it for years, almost completely without mishap. The Fast forward to today . . . Historical Society of Wisconsin, for instance, Lending physical items from special collections has since the early 1970s operated a statewide for research purposes is finally an idea whose network of regional research centers which moves time has come—for some. It remains among the archival materials around so that researchers most divisive issues in the field of archives and can use them close to where they live (Erney special collections, perhaps the one most likely and Ham 1972). More recently, in 2010, Elaine to bring out equal parts raw emotion and well- Engst of Cornell University sent an entire archival reasoned professional opinion. But solid evidence collection to Columbia University so that a indicates that the practice of lending physical Manhattan-based FBI agent could, over many items from special collections is becoming more months, search for crucial provenance evidence commonplace than not doing so. The Sharing in an effort to recover letters allegedly stolen Special Collections Working Group’s 2010 survey of from a special collection at the New York Public 88 special collections and archives departments Library (2012). NYPL had no item-level description in North America, Europe, Australia, and Africa, of the collection, but decades ago a Cornell Ph.D. found that 57.4% of respondents will lend physical candidate consulted it and extensively described items from their special collections within a many of the letters in notes made while preparing consortium, while another 10.3% will lend even his dissertation. Those notes were the key to the beyond their favored group (See figure 1.). That’s case. This represents a classic instance where 67.7% of respondents who lend physical items only prolonged access to a complete set of from special collections at least some of the time. original archival material at a spot near the user’s home base could adequately satisfy the need. But emotions on this issue can run high, and professional peer pressure can be intense. I offer one example from my own experience:

117 Making Archival and Special Collections More Accessible

Figure 1. Most respondents (67.7%) physically lend special collections items

118 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Getting to “Yes” The working group was made up mostly of teams of sharing special collections and a of special collections curators and interlibrary panel discussion featuring grizzled loan supervisors and included seasoned veterans veterans alongside newcomers to the of lending special collections, those who had practice (Schaffner and Massie 2009). experimented with the practice, and one team • Supporting and informing the work of the considering doing so for the first time. One of RBMS Task Force that, in 2011, revised our initial tasks was to develop a set of “first the ACRL guidelines on sharing special principles” to guide our exploration of the issues: collections for exhibit and for research, • Lending a physical item from special with our main contribution being to ensure collections is an exception, appropriate that sufficient numbers of interlibrary loan only when providing a surrogate copy professionals and archivists commented on would fail to satisfy the request. the draft guidelines (See this report’s list of references on page 39 for a link to the revised • Considering a loan from special guidelines, which have since been endorsed collections often requires a flip in by the Association of College and Research mindset from “Why?” to “Why not?” Libraries’ board of directors and the Society • Not everything held in special of American Archivists Council). (ALA 2012) collections is equally special. The most significant activity of the working • Not every requester of special group was creating a set of tools that will help collections material realizes that the institutions reconsider and streamline their item is held in special collections. processes for handling loan requests for special • Let those who are best positioned collections materials. These tools include: to do something do it. • A tiered approach to streamlining • Lending physical items from special workflows associated with lending special collections requires trust, both collections, outlining minimal, moderate internally and externally. and maximum amounts of effort and overhead, to be invoked based on • An interlibrary loan (ILL) of special collections material counts as use. —— the material • Borrowers of special collections should give —— the request serious consideration to being lenders; lenders —— the risk tolerance of curators of special collections should be entitled to and administrators some expectation of success in borrowing. • A model written policy on The working group devoted time and energy sharing special collections to a number of activities designed to promote • A “trust” checklist to serve as a conversation the physical lending of special collections: starter between a prospective lender and • Compiling a glossary for use by the an institution interested in borrowing working and advisory groups (the main an item from special collections contribution of which was to establish This report contains a complete description of that by “special collections” we meant the working group’s activities, plus all of the any material held in formal special tools listed above, and advice on how best to collections or archives departments). use them. Let the sharing begin. And continue. • Conducting a survey (sent via international discussion lists) of current practices and attitudes regarding the sharing of special collections, targeting both special collections and interlibrary loan practitioners. • Producing a webinar, Treasures on Trucks, which featured a recent history

119 Making Archival and Special Collections More Accessible

Highlights of the 2010 Sharing Special Collections Working Group Survey

In April 2010, the working group conducted Major revelations included: a survey in order to solicit current attitudes, • Lending physical items from special practices, policies, and priorities regarding collections is now more common (67.7%) the lending of special collections materials than not doing so, at least within consortia. for research purposes. We cast a wide net, announcing the survey on major primary • Digitizing on demand has become routine. sources and interlibrary loan discussion • Condition of the item is still the lists and inviting any library with a special key to the lending decision. collections department to reply. Survey instructions encouraged respondents to have • Attitudes toward unpublished materials are ILL and special collections staff members more restrictive than toward published. work together in answering the questions. • 36% indicated they have written policies We received 88 responses. Types of responding for sharing special collections (but no institutions varied greatly and included academic, one had an overall policy; each example national, and public libraries, plus museums covered a particular aspect or format). and historical societies. The overwhelming • “Too risky” (69%) is by far the most common majority of responses came from North America, reason for not sharing returnable special with four from continental Europe and one collections (i.e., original items held in each from Africa and Australia. Respondent special collections that must be returned job titles included a mix of special collections at the conclusion of the loan period). curators, reference or access heads, and interlibrary loan supervisors, along with a few • “Because we never have” and “Not part of our university archivists and library directors. mission” each got more votes than “Lack of staff resources” as main reasons not to lend returnable special collections materials. • Most interesting comment: “We were able to borrow things we would not be able to provide to others.”

120 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Figure 2. Issues involved in physically lending special collections items

The first section of the survey focused on spirit of considering each request on its own policy. It was impossible to find any sort of merits—that it deserves to be quoted in full: consensus in the responses, other than one We look at WorldCat to see how many other sizable camp being willing to lend even its libraries have the item and where they are treasures to trusted partners, while another located. If the requester lives within a day’s to- sizable camp expresses an aversion to risk and-fro driving distance, we would usually prefer that at times sounds more like fear. that the researcher come to us to use the book. One respondent (an ILL staff person) expressed If we find via WorldCat that the requester lives surprise at learning through completing the nearby another library (a researcher from the survey that special collections staff often University of Chicago who wants a book that is receive and fill requests directly without ILL owned by the Newberry Library, for example) involvement; the ILL’er felt that such requests we would usually decline to lend, particularly should be routed through the established if the title is scarce. We think about the rigors resource sharing channels, because a of traveling and how the journey might affect willingness to lend such items creates a valuable the condition of the book. Some items are reserve of good will for that library out in the just too frail to lend and must be used under community when it comes time to borrow. curatorial supervision. We think about the type of book—novelty books with pull tabs, fragile In response to an open-ended question about pop-up books, etc.—and turn down requests how the decision is made to lend or not to to borrow. When we are reluctant to lend, we lend, we mostly received confirmation of often look up researchers in their university what was learned from the multiple-choice directory and email them to ask what specifically questions: condition matters most, with other they are looking for. Sometimes we can fill a factors such as rarity, value, popularity, and request by photocopying the table of contents proximity (of the requester to the supplier, or or the index, or perhaps a few relevant pages. of the requester to other copies of the same When requesters learn that the book is in a material) carrying significant weight. But one special collections library, we find that some response was so thoughtful and comprehensive say, “Oh, never mind. It isn’t terribly important” in approach—while so perfectly capturing the

121 Making Archival and Special Collections More Accessible or “I wouldn’t want you to lend a rare book.” mentioned that their institutions do not lend Sometimes we find that researchers have gone special collections and have no plans to review on to another topic or no longer need the book their policies; others lauded the increasing because of a deadline. It sometimes makes a emphasis on access; a few wanted to hear more difference whether a researcher is engaged in an about the experiences of those institutions that initial fishing expedition on a topic or whether are successfully lending entire archives. One the book would contribute to a major project. respondent pushed the idea of digitizing as If we learn that our book is a vital part of a much as possible and making it available online researcher’s work, then we will go out of our way as the best means of providing access. Another to accommodate the request. For books that are wrote, “We receive for our patrons materials that not particularly valuable or scarce, we check are similar to items we would not provide.” in other online catalogs to see if the same title In other words, the survey showed us might be in the circulating stacks of another what we suspected already: that there library. We have lots of books (science fiction, is currently no consensus on any aspect utopias, or works by certain authors, for example) of sharing special collections. that are not particularly rare but that in OUR library are housed in Special Collections because of their subject or provenance. We wouldn’t lend Survey Implications one of our utopian works if circulating copies are easily available from other institutions. We When community practice is all over the map think of what it might mean to us if our book or split down the middle, the time is ripe is damaged or lost. (We have on occasion lost for someone with a strong point of view to books through lending.) There’s a cost involved step forward and lay out a prospective path in two senses: What would it literally cost us for that community. The Sharing Special to replace a book (if we could)? What would be Collections Working Group studied the survey the cost to our collections and researchers if we results and decided to leap into the void. couldn’t replace a book? If we are asked for a particularly scarce book, I look to see if there is a copy currently on the market. (This is often Taking the “Scare” Out how we establish values as well.) Could it be easily replaced? Was it given by a donor of Lending the Scarce who would be angered by our having loaned it? For members of our working group, volunteering Is it a key item within Special Collections that for this assignment meant continually having we couldn’t afford to lose because it is so closely to confront their own fears about the physical identified with us? Is the book unique (a signed lending of special collections materials for copy or an association copy, for example), or research purposes. (See appendix 1, a case does it have a particularly fine binding? We study of Pennsylvania State University staff think about our local use patterns. We have an involvement in this process, for an explicit unwritten policy not to lend county histories example.) These fears became more manageable or county atlases, for example, because they as we built up a core set of working principles. are so often consulted in our own reading room. It would be a hardship to our users (particularly genealogists on the road) to come here to find that a book that is supposed to be non-circulating is at another institution. Principal #1: The appropriate I hope someone thinking like this will answer is still usually “No.” be processing my own ILL request for No one is going to lend The Book of Kells— special collections materials. except perhaps for the most major exhibitions. In all cases, making a surrogate of the item, The survey closed with the open-ended question, digital or otherwise, will be the first option “Is there anything you’d like to tell us about in answer to an external request to borrow sharing special collections materials that wasn’t something from special collections. But will addressed by the survey, or any point you’d like a surrogate be useful to the researcher? to emphasize?” We received 22 responses. A few

122 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Principle #2: Not all special collections A Tiered Approach material is equally special. Many items are in special collections not because With these principles in mind, working group they are rare or valuable, but because they were members set about creating a flexible system written by a certain author or type of author, for considering loans of special collections (See about a certain place or topic, in a specific genre, figure 3.). A flexible approach acknowledges or at a certain time. Some special collections differences in user needs, collections, institutions, items, such as transcripts of oral history tapes, and resources. As always, institutions will bring are easily replaceable and even easier to copy. to bear professional judgment regarding when to scale up effort and investment. Delivery of special collections material, whether of Principle #3: Not all requests for special the actual item or a surrogate, is the goal, no collections material are created equal. matter the combination of tiers chosen. Often a borrower doesn’t realize that a requested item is held in special collections. Sometimes We borrowed the concept of three tiers, or three the researcher really needs to see the original; in levels of effort and overhead, from the work other cases, a copy of only part of the material presented by Jennifer Schaffner, Francine Snyder, will suffice. A researcher may be under a crippling and Shannon Supple in their April 2011 OCLC deadline or may have all the time in the world. Research report, Scan and Deliver: Managing Sometimes the requested item is absolutely User-Initiated Digitization in Special Collections and critical; at other times the researcher is merely Archives. We listed the main steps in processing satisfying an idle bit of curiosity and wouldn’t external requests for research loans of special want to put anyone to any special trouble. collections: review, decide, lend, and return. Next we laid out tiers with three distinct levels Principle #4: Interlibrary loan staff knows how of effort and overhead that may be chosen and combined based on decisions about the to lend things and get them back safely. value, condition, rarity, format, rights status or It’s what they do. They are meticulous. They’ve popularity of the requested item; the identity, spent decades perfecting infrastructure and location, and controlled environment of the techniques. They’re aware that existing national borrowing institution; the status, needs, and and international ILL codes serve as implied point in the research process of the researcher; contracts that cover any and all interlending and the policies, staff capabilities, and available transactions. They know how to double back resources of both institutions. Knowing what to the requester and find out exactly what questions to ask and which level of staff to involve is needed. They established the community at each stage of the process are important first practice of sending out a surrogate instead of steps in streamlining processes, establishing lending the original. They’re experienced in effective communication among cooperating making sure material is handled properly. With departments, and ensuring appropriate a little coaching, they can expertly handle even handling for materials regarded as “special.” the rarest or most fragile material. They will be judicious about when it’s time to confer with special collections experts. In short, special collections and archives staff can trust them.

Routine Workflow Cooperative Workflow Exceptional Workflow REVIEW

Request Via ILL system Collaboration between Directly to SC Special Collections (SC) and ILL

123 Making Archival and Special Collections More Accessible

Routine Workflow Cooperative Workflow Exceptional Workflow Is material held in a ILL staff Collaboration between Lending institution special collection? borrowing and lending institutions

Reference Interview At borrowing Collaboration of By lending institution—reference ILL and SC staff in institution—SC staff desk and ILL staff both institutions

Inter-institutional ILL system ILL system and Direct contact communication how? email/phone between two SC’s

Internal ILL system ILL system and Direct contact between communication how? email/phone SC/ILL staff and other departments

Stipulate for Implicit Consider emphasizing Explicit criteria Research Use?

Reviewing Written guidelines Collaboration between Elaborate decision Infrastructure borrowing and lending tree, multiple departments staff, institutional level decision

Mutual disclosure of We trust you Approved checklist Facilities report ILL and SC facilities

Forms ILL transaction work Extra insurance Use agreement, form and IFM and/or forms for insurance forms, special handling art museums loan agreement, etc.

DECIDE

Decision Maker ILL staff ILL and SC consult SC staff, , when necessary possibly director

Original or Surrogate? Surrogate or Prefer to lend Case-by-case predetermined surrogate, consider consideration originals original

Published/ Some published Some published OK. Consider lending unpublished? and predetermined Unpublished material published and unpublished on a case-by-case basis unpublished materials material types

Use Rights Borrower’s What any reasonable Search, monitor and responsibility SC staffer would do control thoroughly

124 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Routine Workflow Cooperative Workflow Exceptional Workflow Trust and Training ILL training and ILL and SC cross- SC training and expertise training on handling experience only fragile materials

LEND

Oversees loan ILL staff Staff in ILL and SC SC specialists transaction

Quality Control Usual packager, Special ILL or SC/preserv staff usual shipper, SC packager prepare special mailroom or ILL supports and deliver with the material

RETURN

Deliver Usual shipper, Expedited shipper, Deliver from SC to SC— with use/handling extra insurance, special call me when you get it conditions handling instructions

Figure 3. Tiered approach to sharing special collections, with varying degrees of effort and staff involvement

125 Making Archival and Special Collections More Accessible

The Way We Were—and the Way We Could Be While the survey results indicate condition of collections originals, the tolerance for risk. the material as the primary consideration when Meet with your colleagues in special collections deciding whether to lend a special collections and interlibrary loan. What classes of material item, discussion among working group members make sense for each track at your shop? For revealed that the dominant factor for determining what material does it make sense to blend which tiers one will use is attitude toward risk. tracks, taking some steps in the Cooperative Workflow and others in the Routine Workflow? Every research request used to be exceptional. What materials push you outside your comfort Longstanding practice for those institutions level? What do you do when that happens? that considered lending items from special Proceed directly to the Exceptional Workflow? collections was to treat each request according to the far-right “exceptional” tier. Most often Remember to breathe. requests were received directly by special Then have the conversation collections staff; indeed, if the ILL office received about the tracks again. a request for an item in special collections, common practice was to respond negatively The purpose of this report is to bring you to tiers. and advise the borrowing institution to contact the special collections department directly. Often multiple staff members consulted about whether to lend the item. Typically special collections staff contacted the borrowing institution to talk about the patron’s needs and the borrowing staff’s ability to handle a loaned special collections item professionally. In some cases, use agreements and special insurance arrangements were required before a special item was be loaned. Preservation staff sometimes contributed special containers and support structures to protect the material while on loan. Surely, working group members reasoned, there must be another kind of workflow appropriate to processing such requests. Surely there must be whole classes of special collections holdings about which an interlibrary loan person could be relied upon to make lending decisions, beyond a blanket negative. Perhaps there could even be middle-ground just beyond the obvious cases that could be decided cooperatively; special collections and interlibrary loan staff could come to an understanding about classes of material where a minimal amount of consultation would be appropriate, not necessarily to the level of bringing in curators or directors every time, and always with an eye toward providing a surrogate rather than lending the actual item whenever a copy would be sufficient. Surely a system could be put in place where the deluxe take-no-chances approach is saved for those few situations that actually require it. Take a look at the tracks in figure 3. Think about the mindset at your institution, the prevailing attitude toward lending special

126 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Tool 1: Model Local Policy for Lending Special Collections Materials for Research Purposes

Working group members were excited when nearly a third of the respondents to our survey reported having developed a written policy statement that guided the sharing of items from their special collections. While no single institution possessed the kind of comprehensive policy statement that we sought, enough survey respondents provided examples of written policies on sharing particular formats that we were able to borrow the language needed to develop a comprehensive model policy statement on sharing special collections. The lion’s share of this work was done by OCLC Research Program Officer Jen Schaffner, and working group member Scott Britton (then at the University of Miami). Our approach was to provide a multiple-choice template that special collections staff could customize for local use, adding and deleting elements to fit local practice.

Lending and Borrowing Special Collections for Research Purposes: Model Local Policy Mission statement [example; add or delete as needed]: The [institution name] Special Collections unit supports an active program of loans from its collections. We take local demand for special collections into consideration when deciding whether or not loan. The benefit of increased public access to its collections is measured against internal programs and the demands of preparation, packing, and transportation, with special consideration to the physical conditions of the work must endure throughout the loan. Accordingly, all loan requests are subject to a formal approval procedure. All requests are [considered.] [considered on merit.] [considered for their contribution to scholarship/human knowledge.] [considered for their public purpose.] [etc.] Formats [add or delete as needed]: • Formats of materials that will be considered for loan include: [microforms], [rare books], [manuscripts], [maps], [archives], and [videos] [etc.]. • Items and collections for loan must be in stable condition that will not be damaged by the move, change of environment, or even supervised handling by the Borrower. • Items that are fragile, expensive or oversized may circulate with special packaging, handling instruction and insurance. Requests [add or delete as needed]: • Inquiries regarding Interlibrary Loan policy and procedures for special collections should be directed to [ILL email or special collections email] or by telephone at [phone number]. • Researchers must channel loan requests through a qualified institution [university or college library, historical society, public library, archives, museum, etc.]. • Preliminary research concerning a request should be carried out well in advance so that the formal request can be made in a timely fashion. • Requests accepted via: [ALA], [OCLC], [fax], [email], and [telephone]. • The preferred requesting method is [ILL system] [link to forms][extraordinary circumstances and forms]. • The institution charges what is charged for ILL, except in extraordinary cases. Any preparation requested by the Borrower or required by the Lender which is at variance with normal practice will attract additional charges. These will be negotiated on a case by case basis. Additional shipping/insurance costs may also be charged. • Unless otherwise specified in writing, all works will be released from and returned to [your mailing information here]. 127 Making Archival and Special Collections More Accessible

Terms and conditions of loan [add or delete as needed]: • No item may be re-loaned by Borrower to a third party. • The borrowing period shall be for [x days or weeks] with a [x days or weeks] renewal period. • Long-term loans will be considered on a case-by-case basis. • A researcher may borrow up to [x] items at one time and may not request additional loans until previously borrowed items have been returned. • Researchers must be in good standing at their home institution. • In the event that there is a local request for the loaned material, it will be recalled. • The work must be stored in a space equipped to protect it from fire, smoke, or flood damage; under 24-hour physical and/or electronic security; and protected from humidity and temperature extremes, excessive light, and from insects, vermin, dirt, or other environmental hazards. • No statement of valuation will be given an item in any manner to individuals or to the general public. • The Loaning institution recognizes that a Borrower may cancel a loan, or other circumstances may prevent the loan from taking place as planned. Once remitted, loan-processing fees are non- refundable, regardless of circumstance. Terms and Conditions of use [add or delete as needed]: • All loaned materials must be used in the Borrowing library, in a reading room monitored by special collections staff. • Staff of the Borrowing institution will ensure that the lender’s regulations for use of [rare books, manuscripts, special collections, photographs and/or archives, etc.] are enforced during the loan period. • Researchers must handle materials gently, taking care in a manner that avoids damage and excessive wear and tear. • Permission for reproduction, including electronic formats, must be obtained from the Loaning institution. Permission may also need to be obtained from the copyright holder, if any. • For specific digitization and publication use questions, please [visit the website] [contact staff]. • Each reproduction must be labeled and credited to the Loaning institution [as specified]. • Some material may not be available for reproduction due to preservation, copyright or other permission restrictions.

128 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Packing, shipping and handling [add or delete as needed]: • Only qualified staff may unpack, handle and repack the work/s. • Any instructions given by the Lending institution regarding unpacking, handling and repacking are to be followed. • The Borrowing institution will keep the packing materials for return shipment, and the work will be repacked using the same protective methods and materials. • The Lending and the Borrowing institutions will ship the materials by a courier with tracking capabilities, such as UPS or Federal Express. • The Borrowing institution may be required to bear costs associated with the shipping of the work/s including crating, packing, transportation, etc., in both directions. • The Borrowing institution is responsible for returning the materials in the same condition as received. • No work may be altered, cleaned, or repaired without prior written permission. • Any damage, deterioration or loss to the work/s must be reported to the Lending institution immediately. The work/s should not be moved or treated until further instruction from the institution unless necessary to prevent further damage. • If irreparable damage or loss occurs at any time, the Borrowing institution must meet all costs of replacement, or appropriate compensation.

129 Making Archival and Special Collections More Accessible

Tool 2: The SHARES Facility Trust Checklist

Staff at institutions that lend physical items from special collections report that, upon receiving a borrowing request from another library, they often pick up the phone and initiate a conversation with the special collections or interlibrary loan practitioner at the borrowing library. In 2011, a SHARES working group compiled a set of core questions that the prospective lender typically asks of the borrower during such a conversation. This work was led by Aimee Lind of the Getty Research Institute. The aim was to establish a set of core criteria that, when met by an institution requesting special collections material, will allow the curator to lend with confidence that the material will be handled safely and professionally. The 2012 SHARES Executive Group agreed that providing a list of such criteria to prospective borrowers and lenders is a valuable first step in promoting the sharing of special collections materials. Working group and SHARES Executive Group members identified these potential use cases for the checklist: • For a borrowing institution to cite compliance in interlibrary loan requests for special collections materials, as an indication to lenders that the material will be handled safely and professionally. • For a borrowing institution to use to convince its own administration that upgrades in facilities and professional competencies are required in order to borrow materials essential to researchers. • For a lending institution to send to a prospective borrowing institution that has requested special collections material through interlibrary loan, to confirm that the borrower has the facilities and competencies necessary to ensure safe handling of the borrowed item. • In cases where the borrowing institution does not meet all the criteria, to use as a “conversation starter” with prospective lenders who may be willing to be flexible or to provide certain classes of material if a subset of the criteria are met.

130 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

SHARES Facility Trust Checklist: Baseline Criteria for Sharing Special Collections Materials

Institution name and address______

______

Contact info for ILL______

Contact info for Special Collections______

1._____ My institution employs staff trained in handling special collections materials.

2._____ My institution maintains a supervised and secure reading room.

3._____ My institution’s supervised reading room is climate-controlled.

4._____ My institution has a locked storage area or vault for housing special materials.

5._____ My institution’s locked storage area or vault is climate-controlled.

6._____ The bags of those leaving my building are inspected, and/or patrons are required to leave bags in a locker before visiting special collections.

7._____ My institution’s special collections area has intrusion detection equipment.

8._____ My institution’s special collections area has a fire detection system.

9._____ My institution’s special collections area has a fire suppression system.

10.____ My institution has insurance covering loss of borrowed materials due to damage or theft.

11.____ My building has a secure mail receiving room.

12.____ Incoming and outgoing special collections materials are received, unpacked, packaged, and shipped by staff trained in handling special collections materials.

131 Making Archival and Special Collections More Accessible

Appendix I: Case Study—Pennsylvania State University

(This account draws heavily upon • How does the lending institution staff materials prepared by Sandra Stelts and determine that the requester actually Barbara Coopey; see list of references at needs the special material? the end of this report for specifics.) • How do we build trust—not only between Who: Sandra Stelts, Curator of Rare borrowing and lending institutions but also Books and Manuscripts between Special Collections and ILL? Barbara Coopey, Assistant Head, Access • What can be loaned under Services; Head, Interlibrary Loan what circumstances? Pennsylvania State University Libraries • What can be digitized and added to the collections for others to access and use? What: Embraced the idea of considering requests for loans of their special collections One almost paralyzing worry was that Penn materials, including unpublished material. State would be overwhelmed with requests for materials held in their Special Collections, When: After attending the 2009 OCLC Research especially items they considered special webinar, Treasures on Trucks and Other because of subject or provenance, but that Taboos: Rethinking the Sharing of Special other institutions would keep in their general Collections, organized by the group collections. A real breakthrough for Penn State that did the work described in this came during an advisory group conference call report. (Schaffner and Massie 2009) when Eleanor Brown, then of Cornell University, Why: To quote Sandra, “We became alternately reported that her ILL department sends a intrigued and alarmed by the suggestion conditional response to all who request special that special collections curators should collections material through interlibrary loan: consider lending more and more “This item is held in our Special Collections. materials—including original archival If you cannot locate this material elsewhere, and manuscript collections. Such loans please try us again.” Once more, quoting on the surface seem contrary to our Sandra, “It’s so simple, and it has helped us perceived mission and have put special to focus on the requests that are unique to collections curators’ desire to protect our institution. We have also asked our own unique material at odds with interlibrary ILL staff to tell us when Penn State is the only loan librarians who want to fulfill location on a request. We know to take those these requests for these materials.” requests particularly seriously and to make How: Joined the OCLC Research Sharing Special every effort to lend or make surrogates.” Collections Working Group, helped to Liberated by the “conditional response” strategy, develop practices to streamline the Penn State staff proceeded to examine the process of sharing special collections workflow between ILL and Rare Books and materials, and then applied these Manuscripts to ensure careful transport of concepts to improve their own workflow. material between the units. They acquired In applying the thinking of the working group to distinctive tubs (See note, figure 4) that both the situation at their home institution, Sandra protected special material while in transit and and Barbara found that the following questions set it apart from other items being moved in particularly resonated with Penn State’s concerns: and out of ILL. (This has led to some instances of “tub envy” from staff of other Special Collections • Collections are for use; how can we share? units; after some quiet negotiations, archival • Does the user know the material materials being handled for ILL purposes is in a special collection? are now permitted to ride in the same tub as Rare Books and Manuscripts materials.) • Should the request go to Special Collections directly or through ILL?

132 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Figure 4. ILL transaction record of Pennsylvania State University lending a manuscript to Columbia University

They reviewed paperwork that accompanies and concerns, such as effectively balancing loaned material, including instructions for the pressing needs of researchers with the shipping, insurance, and safe handling. They library’s imperative to protect the material. increased the number of filled requests by Barbara recently had a query from a in scanning with an overhead scanner to protect Japan who wanted to know what sort of security fragile material. They made paper “preservation” the Penn State library offered in the reading room copies of fragile items under the copyright before deciding to lend them a book. Barbara just law’s fair use provisions and lent the copy. They happened to have photos taken for a presentation improved measures to ensure the safety of about the Sharing Special Collections working room-use-only materials borrowed from other group and was able to document the layout institutions—as well as their own materials—by and security regime of the reading room. The moving the photocopier to a location directly librarian in Japan loaned the book. Working next to the reference desk to ensure more group members agreed that having such photos direct staff supervision and compliance with on hand to share discreetly during the course no-photocopying rules. Other renovations to of an ILL transaction would be quite useful. the reference and reading rooms will improve sight lines from the reference desk, and the Once involved with lending special collections security cameras have been upgraded. materials via ILL, Penn State staff warmed to the task. They discovered early on that it Sandra and Barbara soon learned that trust was often useful to be in touch directly with was the key ingredient in the sharing of special the other library’s patron to find out exactly collections material; as Barbara put it, “Trust what was required, and how vital the need should exist not only between borrowing actually was. Sometimes they could satisfy a and lending institutions but also between ILL researcher’s request by simply photocopying and Special Collections.” Penn State staff put a table of contents, or a single chapter, or an major effort into building trust between the illustration, rather than lending the whole ILL and Special Collections units by increasing book. Staff discovered that researchers were communication and paying more attention to often sensitive to the curators’ concerns, saying the process. Both units now better understand “Oh, never mind, I wouldn’t want you to ship the concerns and needs of the other and, in fact, a rare book” or “I can try to find it on my next find that they share many of the same needs 133 Making Archival and Special Collections More Accessible trip to Italy—let’s hold off for now.” When the need was truly urgent and could not be satisfied from other sources, Penn State staff went to great lengths to find a way to fill it. All of this work building upon the accomplishments of the Sharing Special Collections Working Group has, in Sandra’s opinion, led to an increased alignment of the Penn State library with the institutional mission. And to quote her one last time: “I bask in praise after a successful transaction, such as ‘Oh! You are just too good!’”

134 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Appendix 2: In-depth Analysis of the 2010 Sharing Special Collections Working Group Survey

In April 2010, the working group conducted materials than for unpublished, while 35.3% a survey in order to solicit current attitudes, do not, and 16.2% “sometimes” have different practices, policies, and priorities regarding policies. Comments revealed that many have the lending of special collections materials the “same” policies for both because they do for research purposes. We cast a wide net, not lend anything from special collections, announcing the survey on major primary published or unpublished. The differences in sources and interlibrary loan discussion policy usually centered around being sometimes lists and inviting any library with a special willing to lend published material but not collections department to reply. Survey unpublished, or to copy published material but instructions encouraged respondents to have not unpublished. There was no consensus. ILL and special collections staff members Only 10.3% of respondents indicated that they work together in answering the questions. lend physical items from special collections to We received 88 responses overall, with 64 other libraries, with another 57.4% reporting completing the entire survey. Types of responding that they will do so “under certain conditions,” institutions varied greatly and included academic, for a total of 67.7% who share physical items at national, and public libraries, plus museums least sometimes. About a third (32.3%) never and historical societies. The overwhelming lend physical items from special collections. majority of responses came from North America, Comments revealed that many lend only for with four from continental Europe and one exhibition, others only to fellow participants in each from Africa and Australia. Respondent the SHARES resource sharing program, still others job titles included a mix of special collections only published materials. One library reported curators, reference or access heads, and experimenting with loans of entire archival interlibrary loan supervisors, along with a few collections to other libraries within their state. university archivists and library directors. Again, there was no consensus on best practice.

ILL Lending Policies for Special Those who do lend physical items from special Collections Materials collections to other libraries were asked to choose the top three issues involved in the decision- The first section of the survey focused on making, from a list of eight that included “Other— policy. It was impossible to find any sort of please specify.” By far the most important issues consensus in the responses, other than one were “condition of item” (noted by 87.2% of sizable camp being willing to lend even its respondents), “rareness of item” (mentioned by treasures to trusted partners, while another 78.7%), and “value of item” (noted by 59.6%). No sizable camp expresses an aversion to risk other answer—age of item, identity of requester, that at times sounds more like fear. location of requester, how busy we are, or Over two dozen respondents (36.8% of the total) other—was chosen by more than a quarter of claimed to have a written policy on lending respondents. “Other” choices put forward special collections. When the working group included how heavily the item is used at the home followed up, however, we found that not a single institution, the value to the home institution institution had an overall written policy covering aside from monetary value, and the quality of the all special collections and archives. Rather, they environmental conditions at the borrowing library. had a written policy on some aspect of sharing, Those who do not lend physical items from their such as microfilms or digitizing out-of-copyright special collections to other libraries were asked materials. In the end, we borrowed language to rank the reasons why they don’t, from a list of from several of these narrowly-focused policies seven that included “Other—please specify.” The to create a model overall policy for sharing most popular reasons were “too risky” (69.2%), special collections materials. (See tool 1.) “other” (51.3%), and “items needed onsite Nearly half of respondents (48.5%) have different (30.8%). “Other” reasons included “items are policies for lending published special collections irreplaceable,” “have loaned previously and gotten 135 Making Archival and Special Collections More Accessible back damaged items,” “resistance on the part of manage library-to-library requests for special special collections staff,” “resistance on the part of collections materials, while 9.5% manage such branch managers,” and, my personal favorite, “an requests in special collections; 33.3% receive and atmosphere of mistrust and fear.” As previously manage such requests in both departments. mentioned, the reasons “not part of our mission” By far the most popular method for managing and “because we never have” each was chosen and tracking ILL requests for special collections twice as often as “lack staff resources,” which materials was ILLiad (41.0%), the ILL management the working group members had anticipated software created by Atlas Systems, with the next being an oft-cited reason for not lending. popular being paper files (16.4%). Other methods Nearly half of respondents (47.8%) reported included spreadsheets, integrated library systems, lending surrogates of special collections materials and Clio, an ILL management package designed to other libraries, while another 35.8% said that by Clio Software. One respondent reported using they do “under certain conditions,” for a total of Aeon, an online request system for archives and 83.6% lending surrogates (compared with 67.7% special collections designed by Atlas Systems. lending physical items). Only 16.4% reported In response to an open-ended question not lending surrogates of special collections about how incoming ILL requests for special materials to other libraries. Comments centered collections materials are “triaged,” answers mostly on the condition of the original item and varied from “we don’t lend” to “we only lend the proportion of the work being requested. within our consortium” to “we check with the One respondent wrote, “We desire to keep our archivist” to “the director reviews the request.” collections, and make our repository valuable The preferred method seemed to be related to to researchers, so we don’t create duplicate the size of the staff handling requests and the collections for storage by other repositories.” volume of requests coming in; busier places Those who do supply surrogates of special saw more of a need to automate and streamline collections materials were asked how they processes; at less busy places or sites where one supplied them, choosing all methods that apply staff member handles all incoming requests, from a list of five, including “Other—please procedures were more informal, epitomized specify.” “Scan and send as file” (81.0%) and by the comment, “When I get a request for “photocopy and provide hard copy” (74.1%) were special collections materials, I set it aside until by far the most popular methods, with “scan, a have a minute to go and see the archivist.” add to own digital collection, and provide a In answer to a question about the preferred link” (41.4%) being the only other choice cited modes of communication among staff processing by more than a quarter of respondents. and reviewing ILL requests for special collections Those who do not supply surrogates of special materials, the most popular were “email” and collections materials were asked why not, with up “face-to-face” (36.1% each). Only 13.1% use the to three reasons to be chosen from a list of seven ILLiad client for such communication. Very few use that included “Other—please specify.” “Risk of paper forms and the telephone. One commenter damage to material” (52.6%) and “Other” (47.4%) emphasized, “The answer is no. Always.” were the only choices selected by more than a When asked an open-ended question about the third of respondents. Comments indicated that effectiveness of current procedures for handling some respondents interpreted the question as incoming ILL requests for special collections being specifically about providing a surrogate of materials, most expressed satisfaction. A the entire special collections item, and they either few suggested that more automation would lacked the resources to do so or felt that such a be helpful, while others noted that key staff request would violate copyright in most cases. outages can throw the system into disarray. Workflows for Managing ILL Requests One respondent (an ILL staff person) expressed for Special Collections Materials surprise at learning through completing the survey that special collections staff often received and The divide in the community about sharing special filled requests directly without ILL involvement. collections materials continued when we looked at workflows for managing incoming requests. More than half (57.1%) have interlibrary loan staff

136 Tiers for Fears: Sensible, Streamlined Sharing of Special Collections

Workflows for Processing ILL Requests to surrogates of special collections items to other Physically Lend Special Collections Materials libraries. As with the decision-making process for lending special collections items themselves, The next section of the survey focused on most respondents (44.8%) indicated that a workflows associated specifically with lending curator makes the final decision about whether physical items from special collections to other a surrogate will be sent (compared with 59.6% libraries. Most respondents (59.6%) indicated having curators decide when loaning the actual that a curator makes the final decision about item); ILL staff make the decision on lending a whether a particular item will be physically surrogate in 25.9% of the responses (compared loaned, with ILL staff making the decision in to only 11.5% having ILL staff decide on lending only 11.5% of the responses; at nearly a third of the actual item). As with the decision-making for the surveyed institutions (28.8%), it is a group lending actual items, nearly a third of the surveyed decision. In response to an open-ended question institutions (29.3%) make providing a surrogate about how the decision is made to lend or not of a special collections item a group decision. to lend, we mostly received confirmation of what was learned from the multiple-choice The survey closed with the open-ended question, questions about basic processing of requests for “Is there anything you’d like to tell us about special collections materials: condition matters sharing special collections materials that wasn’t most, with other factors such as rarity, value, addressed by the survey, or any point you’d like popularity, and proximity (of the requester to the to emphasize?” We received 22 responses, mostly supplier, or of the requester to other copies of reaffirming points made elsewhere in the survey. the same material) carrying significant weight. A few mentioned that their institutions do not lend special collections and have no plans to Most potential lenders of physical items from review their policies; others lauded the increasing special collections don’t require any specific emphasis on access; a few wanted to hear more knowledge ahead of time about the borrowing about the experiences of those institutions that patron, with a few respondents asking to know are successfully lending entire archives. One the name and/or patron status. In response to respondent pushed the idea of digitizing as an open-ended question about what potential much as possible and making it available online lenders of such material might want to know as the best means of providing access. Another ahead of time about the borrowing institutions, wrote, “We receive for our patrons materials that most mentioned the security and environmental are similar to items we would not provide.” controls in place, or the presence of professional staff to supervise use of the items. A few would The survey showed us what we suspected want to know if the borrowing institution was a already: that there is currently no consensus fellow member of a consortium such as SHARES. on any aspect of sharing special collections. A very few indicated that they would not lend When community practice is all over the map special collections materials to a public library. or split down the middle, the time is ripe As for packaging special collections materials for someone with a strong point of view to for loans to other libraries, respondents were step forward and lay out a prospective path almost evenly split between assigning this task for that community. The Sharing Special to ILL staff (34.0%) and special collections staff Collections Working Group studied the survey (30.0%). Only 10.0% of respondents delegated results and decided to leap into the void. such packaging to the mail room. “Other, please specify” responses comprised more than a quarter of the total (26.0%); they varied from a division of labor (conservator makes special boxes, ILL staff does packaging) to a case-by- case approach based on condition or format.

Workflows for Processing ILL Requests for Surrogates of Special Collections Materials The final section of the survey focused on workflows associated specifically with lending

137 Making Archival and Special Collections More Accessible

References

ALA (American Libraries Association). 2012. “ACRL/RBMS Guidelines for Interlibrary and Exhibition Loan of Special Collections Materials.” Approved by the ACRL Board of Directors January 2012. http://www.ala. org/acrl/standards/specialcollections. Coopey, Barbara, and Sandra Stelts. 2011. “SHARES Special Collections Workflow Group and Penn State’s Efforts to Share.” Presentation at the Association of College and Research Libraries Conference, Philadelphia, Pennsylvania, April 2011. http://www.learningtimes.net/ acrl/2011/when-worlds-collide-interlibrary- loan-and-special-collections/ (presentation starts at about the six minute mark) Engst, Elaine. 2012. “Baseball, True Crime, the FBI and I(LL): Interlibrary Loan for Archival Collections Revisited.” Presentation at the Society of American Archivists Annual Meeting, San Diego, California, August 2012. http://files.archivists. org/conference/sandiego2012/508-Engst.pdf Erney, Richard A., and F. Gerald Ham. 1972. “Wisconsin’s Area Research Centers.” American Libraries. (February):135-140. Schaffner, Jennifer, and Dennis Massie. 2009. Treasures on Trucks and Other Taboos: Rethinking the Sharing of Special Collections. Webinar produced by OCLC Research. 5 June. http://www5. oclc.org/downloads/programsandresearch/ parcasts/20090528Schaffner&Massie.mp4. Schaffner, Jennifer, Francine Snyder, and Shannon Supple. 2011. Scan and Deliver: Managing User-initiated Digitization in Special Collections and Archives. Dublin, OH: OCLC Research. http://www.oclc.org/research/ publications/library/2011/2011-05.pdf. Stelts, Sandra, and Barbara Coopey. 2012. “ARL Submission For Special Issue of Research Library Issues (RLI).” Unpublished proposal to submit a written account of Pennsylvania State University’s efforts to share special collections materials, to the Association of Research Library’s journal RLI.

138

Making Archival Makingand Archival Accessible More Special Collections

Making Archival and Special Collections More Accessible OCLC Research

1504/215371, OCLC ISBN:-55653 1 3 -- 497 3 - -497 55653 - 1 978-

T: 1-800-848-5878 T: -6000 764 - +1-614 T: 764-6096 - F: +1-614 www.oclc.org/research Kilgour Place 6565 Kilgour -3395 43017 Ohio Dublin, www.oclc.org/research/themes/research-collections.html For more information about our work related to to related work our about information more For visit: support, please and collections research